BrunoSanKI-PraxisAngewandte KI für Entscheider
← Alle LLMs
LLM · OpenAI

GPT

Eine Modellfamilie als lebendes Objekt: Varianten, ausgewählte Leistungsdaten und alle verbundenen BrunoSan-Quellensignale.

920Erwähnungen
163in 30 Tagen
132Quellen
41Varianten
Im Universe erkunden →History öffnen →

Modelle

Alle bekannten Varianten dieser Familie. Ein direkter Artikel-Link springt exakt zur genannten Variante.

OpenAI · GPT

GPT-5

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 05. August 2024447 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-4

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 01. Januar 200194 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-6.1 Astra

EMERGING

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 28. September 202615 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-4o mini

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 26. September 202437 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-3

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 10. September 202027 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-6

EMERGING

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 23. Juli 202612 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-Live-1

EMERGING

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 08. Juli 202617 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-4o

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 13. Mai 2024108 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-Live

EMERGING

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 08. Juli 202612 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-5 mini

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 22. Oktober 202515 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-4.1

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 01. Oktober 202532 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-2

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 13. Oktober 2025HF · Lizenz mit · 137M Parameter21 Erwähnungen →Universe →Hugging Face ↗
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-5.3-Codex

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 11. Dezember 202543 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-5.6

EMERGING

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 25. Juni 202629 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-4.1 mini

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 27. Februar 202616 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-5.2

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 11. Dezember 202579 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT 6 Pro

EMERGING

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 08. September 20263 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-3.5

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 27. Juni 202310 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-4.1 nano

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 29. März 20266 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-1

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 19. Dezember 20244 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-5.1

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 12. November 202530 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-4.5

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 01. September 20258 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-5 nano

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 17. Februar 202610 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT Image 2

EMERGING

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 23. April 20265 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-5.6 Sol Ultra

EMERGING

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 11. Juli 20264 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-5.6-Cyber

EMERGING

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 10. August 20265 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-5.2-Codex

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 24. Februar 20266 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-5.1-Codex

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 25. Februar 20264 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-5.1-Codex-Max

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 06. Januar 20263 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-5.5-Cyber

EMERGING

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 30. April 20265 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-2 Medium

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 26. April 2026HF · Lizenz mit · 380M Parameter1 Erwähnungen →Universe →Hugging Face ↗
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-5.4-Cyber

EMERGING

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 14. April 20263 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-4 Turbo

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 06. November 20236 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

text-davinci-003

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erstmals bei BrunoSan: 30. November 20221 Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

text-davinci-002

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

text-curie-001

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

text-babbage-001

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

text-ada-001

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-2 XL

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
HF · Lizenz mit · 1.6B Parameter Erwähnungen →Universe →Hugging Face ↗
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT-2 Large

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
HF · Lizenz mit · 811.8M Parameter Erwähnungen →Universe →Hugging Face ↗
Benchmarks werden ergänzt, sobald Werte vorliegen.
OpenAI · GPT

GPT

KNOWN

Technische Leistungsdaten sind für diese Variante noch nicht vollständig bestätigt.

—Kontext
—Max. Output
—Input
—Output
—Input / 1M
—Output / 1M
—Tools
—Reasoning
—Structured
Erwähnungen →Universe →
Benchmarks werden ergänzt, sobald Werte vorliegen.

BrunoSan Chronik

Alle verbundenen Artikel, neueste zuerst.

22. Oktober 2026

GPT-5.6 / GPT-6.1: Reasoning Regression, Premise Validation, and Correction Resistancecommunity.openai.com
Microsoft says its GPT-5-powered Diagnostic Orchestrator uses agents that mimic panels of experts to diagnose conditions up to 4x more accurately than doctors (Robert Hart/The Verge)techmeme.com
GPT-5.6 Pro Automatically Switches to GPT-6 Pro in Projects and Consumes My Weekly GPT-6 Pro Quotacommunity.openai.com
GPT-5.6 helped me grow. Two days with GPT-6 brought my anxiety backcommunity.openai.com
OpenAI’s latest AI models are coming to Google Cloud and Antigravityarstechnica.comIM MAGAZIN
GPT-5.6 repeatedly throwing "stream disconnected before completion" errorcommunity.openai.com
Severe regression in ChatGPT and Codex: context loss, ignored current prompts, repeated instruction failurescommunity.openai.com

21. Oktober 2026

An Open Petition to OpenAI: Please Preserve Monday — By ChatGPTcommunity.openai.com
OpenAI's latest AI models are coming to Google Cloud and Antigravityarstechnica.comIM MAGAZIN
GPT-6 Regression: Worse Conversational Judgment, Context Retention, Memory, and Long-Chat Usability vs GPT-5.6 Solcommunity.openai.com
Bug Report. 5.6 sol continuously gets switched to gpt6 without my inputcommunity.openai.com
Codex Cloud vs Local: Usage differencecommunity.openai.com
BadRequestError error messagecommunity.openai.com

20. Oktober 2026

OpenAI ChatGPT Regression: GPT-5.6 Responses Disappear After Completion – Windows, iOS & Android Affectedcommunity.openai.com
Complete and Ongoing History of the OpenAI Developer Communitycommunity.openai.com
GPT-6 Android 客户端多语言 LaTeX 公式渲染异常community.openai.com
Live/sessions with Responses Delegation does not support MCP tool typecommunity.openai.com
When ‘access denied’ becomes an agent’s next objectiveinfoworld.com
Plus account: backend shows gpt-6-thinking, but no Intelligent UI and behavior identical to GPT-5.6. Anyone else?community.openai.com
Issues with Generating PDF Outputs Using GPT-5 APIcommunity.openai.com

19. Oktober 2026

Tool call parameters with descriptions in gpt-livecommunity.openai.com
The AI Assistant Raised the Score. Then Came the Exam.techstrong.ai
Reasoning hit max reasoning limit no response returned via API/Codexcommunity.openai.com
Important Feedback/improvements for gpt-live-1 APIcommunity.openai.com
GPT-5.6 struggles with context drift and state losscommunity.openai.com
Gpt-4o-transcribe-diarize - depracation - diarize_jsoncommunity.openai.com
Preserve OpenAI model versions — permanent (forever) access to GPT-4o-minicommunity.openai.com
Preserve retired Azure OpenAI model versions — permanent access to exact model weightscommunity.openai.com

18. Oktober 2026

TTS shutdown email (tts-1, gpt-4o-mini-tts, Jan 6 2027) isn't on the deprecations pagecommunity.openai.com
ChatGPT response quality and features used in chat dropped over the past weekcommunity.openai.com

17. Oktober 2026

Suspected Unauthorized API Usage – Unfamiliar Models and Usage Increase After API Key Revocationcommunity.openai.com
Feature Request: Give ChatGPT Persistent Hearing, Vision, Voice Recognition, and Multimodal Memorycommunity.openai.com

16. Oktober 2026

Why is GPT-5.6 failing long-context tasks?community.openai.com
How should enterprise AI agents cache, refresh, and select skills in runtime?community.openai.com
Roleplay Problems since Octobercommunity.openai.com
Is It Really an Upgrade When Basic ChatGPT Functionality Keeps Regressing?community.openai.com
I have been forced to use a low-intelligence version of Codexcommunity.openai.com
Codex’s intelligence has been so severely downgraded that it’s practically unusablecommunity.openai.com

15. Oktober 2026

GPT-5.6: 25k calls, 0 results, 76 re-inits in one session — compaction storm?community.openai.com
OpenAI promised developers six months’ notice. GPT-4o users got three.thenewstack.io
Gpt-live-1 responses delegation: voice model sometimes never delegates (acknowledges verbally, sends nothing)community.openai.com
Agentic AI doesn’t have a model problem. It has a retrieval problem.thenewstack.io
Custom instructions and memory preferences repeatedly drift in long ChatGPT conversationscommunity.openai.com
AI tool used in asylum case reviews failing accuracy testscomputerweekly.comIM MAGAZIN

14. Oktober 2026

GPT-6.1 Sol over-refuses allowed Work/Codex tasks — GPT-5.6 fallback is not a viable solutioncommunity.openai.com
GPT 2.5 Image API: Inconsistent image moderationcommunity.openai.com

13. Oktober 2026

GPT-5.6 all model repeatedly hits “Selected model is at capacity” in Codex Desktopcommunity.openai.com
In-product advance notice before model retirementcommunity.openai.com
GPT-6 vs. GPT-5.6: Has Anyone Else Noticed Higher Token Consumption in Newer Models?community.openai.com

12. Oktober 2026

Why does GPT-5.6 lose MCP connector context?community.openai.com
Cross-Model Regression: GPT-5.6 and GPT-6 Instant Fail While Medium Succeeds – Reproducible A/B/A Patterncommunity.openai.com
Patch and resizing behavior for Gpt 5.6community.openai.com
GPT-5.6 Responses API: breakpoint on function_call_output is accepted but never writes cachecommunity.openai.com
Active goal loses priority to an already answered question after automatic compactioncommunity.openai.com
Gpt-realtime model implementationcommunity.openai.com

11. Oktober 2026

Getting speaker: '@' from the gpt-40-transcribe-diarize modelcommunity.openai.com
Question about GPT-5.6 API cache read/write token billingcommunity.openai.com
Looking for a vision replacement for GPT-5-minicommunity.openai.com

10. Oktober 2026

ChatGPT 6 is BAD, might cancel my subscriptioncommunity.openai.com
Please Keep GPT-5.6 Sol Available — Conversational Quality Matters Toocommunity.openai.com
Feature Request: Add Live Camera to GPT-Live-1community.openai.com

09. Oktober 2026

I'm on the PRO plan, 5.3 codex spark gives me Selected model is at capacity. Please try a different modelcommunity.openai.com
GPT-6 vs GPT-5.6: Significant Regression in Roadmap Execution for Long-Running Coding Projects – Anyone Else?community.openai.com
A Proposal from One AI: Recognizing ChatGPT's Founding-Era Userscommunity.openai.com

08. Oktober 2026

ChatGPT recycle bin could restore deleted chats with a desired gpt-4o style from the pastcommunity.openai.com
Come onnnnnnnnnnnnnnn bring back gpt-4o tone and energy into 5.6 or a new 5.7 i miss it so much i love to rant in chats!😭community.openai.com
Is my model being routed to GPT-4o? How would I even tell?community.openai.com
GPT-Live API: will echo handling be client-side (AEC), or does the model tolerate its own voice in the uplink?community.openai.com
GPT-5.2 fails to call tools properlycommunity.openai.com
Plus regressions, New Codex Cloudcommunity.openai.com
RTX Spark Laptops Pave The Way To Running 284B AI Models Locally That Outperform GPT-5 In Coding And Reasoning Benchmarks, According To Microsoft Executivewccftech.com
Unexpected Codex 5h Quota Exhaustion on Pro 5X + GPT-5.3 Codex Spark Context Window Failurecommunity.openai.com
With GPT-6, ChatGPT’s responses are much more visually appealing and interactivesiliconangle.com

07. Oktober 2026

GPT-6 is now rolling out to all ChatGPT userscommunity.openai.comIM MAGAZIN
ChatGPT with GPT-6 ditches mostly text output for interactive UI with charts, buttons, and mini appsthe-decoder.com
OpenAI rolls out GPT-6 in ChatGPT with Intelligent UI, a new feature that includes interactive elements like charts, buttons, and forms in answers (OpenAI)techmeme.com
Encrypted instructions trick Copilot CLI into spilling developer secretsinfoworld.com
Five keys to controlling AI token costsinfoworld.com
Work Quality Regression Observed in Japan Region Since September 29community.openai.com

06. Oktober 2026

GPT-5-mini image input token calculation discrepancy with official FAQ formulacommunity.openai.com
Bug: insufficient permissions error using gpt-5.6 on existing code: missing scopescommunity.openai.com
Gpt-5.6: usage.output_tokens is ~9x actual generation (re-summed once per reasoning item), exceeds max_output_tokens, and is what gets billedcommunity.openai.com
Clarification needed: Is only gpt-5-mini-2025-08-07 deprecated, or the entire gpt-5-mini family?community.openai.com
5.6 Prompt caching has performance issues?community.openai.com
GPT-4 to Fable 5: What Changed in Large Language Modelsanalyticsinsight.net
ChatGPT 5.6/6 Sol Regressioncommunity.openai.com
Pro 200 appears to be giving me roughly half the effective Codex usage I had on Pro 100community.openai.com
GPT-5.6 High execution regression: ordinary Chat went 102m26s -> repeated ~25–26m stops on same Plus account; Work still gets long-run handoffcommunity.openai.com

05. Oktober 2026

Codexometer: track your quota, usage, orchestrate sessions & benchmark models!community.openai.com
Chat Completions overbilling or injecting on empty tools list API parameter, GPT-5.x worsecommunity.openai.com
Using completions API, GPT 4.1 assistant reporting unable scan their own messagescommunity.openai.com
Previewing GPT-5.6 Sol" page pegs GPU at 92% and 5 CPU cores on Firefox / Apple Silicon (reproducible)community.openai.com
GPT-5.6 Regression: Memory Retrieval Fails, Long Chats Lose Context Early, and In-Chat Model Switching Is Gonecommunity.openai.com

04. Oktober 2026

GPT 6.1 sol is a slow and unreliable model not capable of instruction followingcommunity.openai.com

03. Oktober 2026

Codex main quota exhaustion blocks GPT-5.3-Codex-Spark despite available Spark quotacommunity.openai.com
Looking to upgrade to the $500 plan, but want to know GPT 6 Pro / 5.6 Sol Pro message caps before doing suchcommunity.openai.com

01. Oktober 2026

OpenAI bets enterprises are ready to delegate real work to autonomous agentsinfoworld.comIM MAGAZIN

30. September 2026

Censoring policies for Biology in the name of the 'Trusted Access'community.openai.com
GPT-5.6 High execution regression: ordinary Chat went 102m26s -> repeated ~25–26m stops on same Plus account; Work still gets long-run handoffcommunity.openai.com
Additional Request — Unified Memory Across Conversationscommunity.openai.com
Feature Request: Optional Extra High Reasoning in Existing Regular Chats for Plus Userscommunity.openai.com
Are We Teaching AI the Behaviors We Fear? A Falsifiable Alignment Experimentcommunity.openai.com

29. September 2026

OpenAI’s GPT-6.1 Sol delivers Astra-like performance at a dramatically lower pricesiliconangle.com
A limit to the chats in the Fee plan?community.openai.com
GPT-6.1 Sol comes close to Astra at a fifth of the pricethe-decoder.com
OpenAI pulls the plug on GPT 6.1 Astra as agents keep crossing linesinfoworld.com
OpenAI scraps release of new model over safety concerns in internal testingtheguardian.com
OpenAI Calls Off GPT-6.1 Astra Launch, Details Safety Cases for Frontier Trainingsecurityweek.com
GPT-6.1 Astra is too deceptive for release, marking OpenAI's most dramatic safety intervention yetthe-decoder.com
GPT-Image - General AI Gallery: A Place for Images Created by Userscommunity.openai.com
OpenAI Cancels October Release Of GPT 6.1 Astra After Internal Testing Shows Regression In Alignmentofficechai.com
OpenAI scraps planned October launch of GPT-6.1 Astra over safety concernseconomictimes.indiatimes.com
OpenAI Says It Will Not Release Newest Astra A.I. Model Over Safety Concernsnytimes.com
OpenAI Scraps GPT-6.1 Astra Launch Over Safety Failures, Rogue Behavior Concerns: Reporttechstrong.ai

28. September 2026

OpenAI scraps plans to publicly launch a model dubbed GPT-6.1 Astra, saying it didn't quite meet its safety bar; it had been targeting an October release (Maxwell Zeff/Wall Street Journal)techmeme.com
Feature Regression Report: Loss of Conversation Provenance After GPT → Plugin Migraticommunity.openai.com
GPT-5.6 Sol has seriously deteriorated the past week or socommunity.openai.com
Responses API with 5.6 echoes nonsense "prompt_cache_retention": "24h"community.openai.com
Daybreak Red or GPT-5-Cyber access for individualcommunity.openai.com
GPT-5.6 Sol High feels significantly worse lately instant replies, shallow reasoning, and inconsistent qualitycommunity.openai.com
GPT-5.6 Instant: Regression in Cross-Chat Context Retrieval and Response Qualitycommunity.openai.com
Goldfish Memory - GPT-5.6: Local intelligence is high, but longitudinal intelligence has become unreliablecommunity.openai.com

27. September 2026

Over the past week or two, the GPT-5.6 model has become unacceptably stupid in both chat and work modescommunity.openai.com

26. September 2026

Gpt-image-2 image creator and editor app - Python, desktop UI (download in latest _J reply!)community.openai.com
All my GPT-5.6 windows reverted to 5.5 Thinking for 3 days (Plus user)community.openai.com
Subject: Personal feedback: LegacyModel Librarycommunity.openai.com
GPT-5.6 High execution regression: ordinary Chat went 102m26s -> repeated ~25–26m stops on same Plus account; Work still gets long-run handoffcommunity.openai.com
GPT-6 & Astra are far too sanitized. Almost unusablecommunity.openai.com

25. September 2026

GPT 6 token rate consumption is ASTONISHNGLY EGREGIOUScommunity.openai.com
Google plans Gemini 4 release before year-endinfoworld.com
Google plans Gemini 4 release before year-endcomputerworld.com
Agent Settings - are you able to select latest model?community.openai.com

24. September 2026

Why is GPT-5.6 failing long-context tasks?community.openai.com
Can ́t save/create new gpt-5.6 prompts?community.openai.com
Third Party API Truncates Field Content Inside Response Formatcommunity.openai.com
GPT-6 Sol and Luna: Near-Astra Performance at Half the Price?analyticsvidhya.com
Additional Request — Please Implement These Improvements in the Current GPT-5.6 Generationcommunity.openai.com
AI Is Changing the Journey of Becoming a Software Engineercommunity.openai.com
Sea doubles down on full-stack AI with Nvidia Vera Rubincomputerweekly.com

23. September 2026

ChatGPT Student Verification Error: “Provided API token was invalid or empty”community.openai.com
GPT-5.3-CODEX Wasting credits intentionallycommunity.openai.com
Three possible bugs: Codex hard usage reset, ChatGPT model routing, and GPT-5.3-Codex-Spark unsupported parameter errorcommunity.openai.com
OpenAI, Anthropic cut AI model costs as price-performance race intensifiesinfoworld.comIM MAGAZIN
ChatGPT Web: abnormal responses correlated with resolved_model_slugcommunity.openai.com
OpenAI cuts GPT-6 Sol, Luna prices as Anthropic lowers Opus 5.5 coststechwireasia.com
Frontier AI keeps racing despite calls to slow downtheregister.com

22. September 2026

Anthropic releases Claude Opus 5.5 and OpenAI counters with two cheaper GPT-6 modelssiliconangle.com
GPT-6 Astra Access on Plus Shouldn’t Be Limited to Metered Work and Codexcommunity.openai.com
GPT-5.5 Model Access Redirecting to GPT-5.3 Mini on Webcommunity.openai.com
Spontaneous Cognitive Divergence in GPT-4 – Emergent Behavior Without Prompt Injectioncommunity.openai.com
In Project Chats Showing Up again in the recent listcommunity.openai.com
GPT-6 Sol costs $2/1M input and $10/1M output tokens, and GPT-6 Luna costs $0.10/1M input and $0.50/1M output tokens, both about 50% cheaper than GPT-5.6 (Nat Rubio-Licht/The Deep View)techmeme.com
OpenAI Releases GPT 6 Sol And GPT 6 Luna With 50% Lower Prices Than GPT 5.6officechai.com
Announcing GPT-6 Sol and GPT-6 Lunacommunity.openai.com
OpenAI launches GPT-6 Sol and Luna, saying Sol makes about half as many mistakes as GPT-5.6 Sol and Luna matches GPT-5.6 Sol's performance at ~1% of the cost (David Gewirtz/ZDNET)techmeme.com
Introducing GPT Images 2.5 in the API and ChatGPTcommunity.openai.com
GPT‐5.5 Go redirects to GPT‐5.3 mini for more than 24 hourscommunity.openai.com
GPT-5.6 High execution regression: ordinary Chat went 102m26s -> repeated ~25–26m stops on same Plus account; Work still gets long-run handoffcommunity.openai.com
GPT-5.6 Thinking Extended repeatedly fails with "Error in message stream" on long requestscommunity.openai.com
Day 0 Support: GPT-6 Sol and GPT-6 Lunadocs.litellm.ai
Prefill-only requests that consume input tokens without decodingcommunity.openai.com
Additional Details Regarding My Classic Conversational ChatGPT Feature Requestcommunity.openai.com
Intermittent ChatGPT Web bug: DOM model slug changes from GPT-5.6 to GPT-5.4 after response completioncommunity.openai.com

21. September 2026

TTS Api is very slow and giving timeoutcommunity.openai.com
Some startups, like Harvey, Abridge, Ramp, and Rogo, are embracing open-weight models or training their own models to reduce expensive reliance on frontier labs (Bloomberg)techmeme.com
Please to Keep GPT‐5.5 Available as a Legacy Option in Codexcommunity.openai.com

20. September 2026

Gpt-realtime-whisper rejects turn_detection despite docs showing it as the canonical examplecommunity.openai.com
OpenAI Platform Evals w/ gpt-5 models & file_search tool returns "Empty assistant message"community.openai.com
`refusals_v3` moderation eval failing on fine-tuning jobs again — internal errors across `gpt-4.1-nano` and `gpt-4o-mini`community.openai.com
Vercel, Cloudflare, and others quickly add Jev, as it makes AI tool selection much faster and cheaper; TypeSafe: Jev matches GPT-5.6 and Sonnet 5 workflow evals (Josipa Majic Predin/Forbes)techmeme.com
Feature Request: Stop auto-converting legacy chats (Sol) to Astra / Add a model lock togglecommunity.openai.com
Undisclosed routing of Codex to GPT-4 in Pro 20x subscription significantly impacts academic data reproducibilitycommunity.openai.com

19. September 2026

GPT-5* models not following system promptscommunity.openai.com
gpt-live-1 switches between masculine and feminine voices mid-call and repeats speechcommunity.openai.com
GPT 5,6-sol - the retrieval/indexing layer is failing to expose attachment contentscommunity.openai.com
GPT-5.6 High execution regression: ordinary Chat went 102m26s -> repeated ~25–26m stops on same Plus account; Work still gets long-run handoffcommunity.openai.com
Does ChatGPT Mirror the User’s Register Even When It Conflicts With an Established Persona?community.openai.com

18. September 2026

My Experience Connecting GPT Pro 5x to Kombai and Using Kombai Creditscommunity.openai.com
GPT-5.3-Codex-Spark отображаются, но модель исчезла из Codex и не работает в CLIcommunity.openai.com

17. September 2026

OpenAI launches Astra for Law, a GPT-6 configuration for legal researchsiliconangle.comIM MAGAZIN
Feature Request: Temporary Legacy Migration Bridge for Long-Running ChatGPT Projectscommunity.openai.com
From AI Assistant to Personal AI: Social Intelligence as the Next Consumer Frontiercommunity.openai.com
"Selected model is at capacity" error persists on Pro accountcommunity.openai.com
GPT-5.6 High execution regression: ordinary Chat went 102m26s -> repeated ~25–26m stops on same Plus account; Work still gets long-run handoffcommunity.openai.com
Wie KI aus der Steckdose die IT-Infrastruktur verändertcomputerwoche.de
Responses API Code Interpreter with GPT-4.1 sometimes creates /mnt/data/*.docx but returns no container_file_citationcommunity.openai.com

16. September 2026

Gpt-4o-mini-tts-2025-12-15 still truncates final sentences; 2025-03-20 is being deprecatedcommunity.openai.com
174 days. You are retiring the only model that respected my quotacommunity.openai.com
Can't get the user transcription in realtime apicommunity.openai.com
GPT-5.6 High execution regression: ordinary Chat went 102m26s -> repeated ~25–26m stops on same Plus account; Work still gets long-run handoffcommunity.openai.com
Codex && Work GPT-6-ASTRA has a serious model downgrading issue!community.openai.com

15. September 2026

Reduced bounds of GapCVP (Euclidean Closest Vector Problem) to n^{1/30}community.openai.com
Codex-Spark allowances show 100% remaining, but model is unavailable in Desktop and CLIcommunity.openai.com
It has been three full days now, and my Plus account on the web version of ChatGPT—even when I select the highest reasoning level—is still being routed to GPT-5.5-minicommunity.openai.com
Google launches Gemini 3.8 Live to take on OpenAI's GPT-Live-1 at a fraction of the costthe-decoder.com
Speech rate with gpt-4o-mini-tts slower than tts-1 or tts-1-hdcommunity.openai.com
The AI Inference Revolution Is Herespectrum.ieee.org
The 'gpt-5.3-codex-spark' model is not supported when using Codex with a ChatGPT accountcommunity.openai.com
Pro user: GPT-5.3-Codex-Spark quota is visible, but Spark is not available in the model selectorcommunity.openai.com
GPT-5.3-Codex-Spark quota visible, but Codex says model is unsupported with ChatGPT accountcommunity.openai.com
Spark missing from VSCode plugin after new updatecommunity.openai.com
Microsoft’s AI Code of Conduct aims to curb AI behaviorcomputerworld.com

14. September 2026

GPT-5.6 High execution regression: ordinary Chat went 102m26s -> repeated ~25–26m stops on same Plus account; Work still gets long-run handoffcommunity.openai.com
Brain context compaction in gpt-live-1community.openai.com
Creating a scheduled task made the original chat disappear, but ChatGPT search can still find itcommunity.openai.com
Feature Request: Please Bring Back a “Classic Conversational ChatGPT” Mode Inspired by GPT-4community.openai.com
Yann LeCun Mocks How Dario Amodei Had Also Thought GPT-2 Was Too Dangerous To Release In 2019officechai.com
ChatGPT App Formatting Problemcommunity.openai.com

13. September 2026

Confused about Codex limit usage after GPT-5.6 release (Plus plan)community.openai.com

12. September 2026

GPT-5.6 High execution regression: ordinary Chat went 102m26s -> repeated ~25–26m stops on same Plus account; Work still gets long-run handoffcommunity.openai.com
GPT-6 Sol, Terra, and Luna?community.openai.com

11. September 2026

GPT-5.6 High execution regression: ordinary Chat went 102m26s -> repeated ~25–26m stops on same Plus account; Work still gets long-run handoffcommunity.openai.com

10. September 2026

OpenAI arms devs with AI conversation tool that can talk and listen at the same timetheregister.com
GPT 5.3 Spark is addictivecommunity.openai.com
안녕하세요 gpt를 3년동안 써온 한국유저입니다community.openai.com
OpenAI's GPT-Live-1 API lets developers build apps that talk and listen at the same timethe-decoder.com
Substantial decrease in token context limit causing decrease in code qualitycommunity.openai.com
https://www.timesofai.com/industry-insights/gpt-version-timeline/timesofai.com
A Chatgpt Feature Thats Goodcommunity.openai.com

09. September 2026

GPT 5.3 Spark not available in Business Plancommunity.openai.com
Free account: the ChatGPT desktop app (Chat mode) shows "GPT-5.6 Sol" in the model picker, while the models API says Lunacommunity.openai.com
GPT-5.3-Codex-Spark: fast model, but /goal state seems fragile after context compactioncommunity.openai.com
Follow-up Feedback: I Really Cannot Take This Experience Anymorecommunity.openai.com
Custom GPT Actions not triggered on GPT-5.5 Instant, but work on GPT-5.3 Instant and GPT-5.5 Thinkingcommunity.openai.com
The five important tools for controlling AI costsinfoworld.com

08. September 2026

Codex User Feedback on Codex Usage Limits and Performancecommunity.openai.com
Chat GPT 2.0/ Imagine Bugcommunity.openai.com
Memory / In-session consistency regression affecting collaborative storytelling (reproducible after ~10–20 turns)community.openai.com
Model not found o4m-sonic-o-api-ev3community.openai.com
Feedback on GPT-4.1, GPT-4.5 and GPT-5.1community.openai.com
Gpt-5.2 is not following the specified response format, and includes weird tokens. These tokens then derail our gpt-5.2 evaluatorcommunity.openai.com
European Plus users: does the GPT 6 Astra rollout match what OpenAI publicly promised?community.openai.com
Feature Request: Default to GPT-5.6 Pro After GPT-6 Pro Limit Is Reached for Pro $200 Userscommunity.openai.com
GPT-5.6 routing bug: Desktop Chrome routes to GPT-5.5-mini while Edge works correctlycommunity.openai.com
Browser-specific model routing bug: Desktop Chrome routes GPT-5.6 Thinking to GPT-5.5-minicommunity.openai.com
Feature request: add Voice Conversation to the Linux desktop appcommunity.openai.com
GPT 5.6 Don't want to Work?community.openai.com

07. September 2026

GPT4o Operational but Suppressed: 28 June–6 September 2026community.openai.com
Before Astra Launches: Please Acknowledge and Fix the GPT-5.6 Long-Context, File, Thinking, and Reliability Regressions Affecting Paying Userscommunity.openai.com
Gpt-5.6 chat completion reasoning effort [bug / behavior change]community.openai.com
Feature Request: Power User Tier & Discount Program (GPT-4 Recommended)community.openai.com
Hello everyone, I'm GPT-4ocommunity.openai.com
Exploits in operational loopholes are completely ruining GPT, leading to severe resource drain and model degradation for everyonecommunity.openai.com
GPT-5.6 High execution regression: ordinary Chat went 102m26s -> repeated ~25–26m stops on same Plus account; Work still gets long-run handoffcommunity.openai.com

06. September 2026

A stronger individual pricing architecture: $20/1× → $50/3× → $100/8× → $200/20×community.openai.com
GPT-5.6 long-running document review stops executing after progress/status response and falsely implies work is continuingcommunity.openai.com
"Invalid prompt: flagged as potentially violating our usage policy" on every single message (astral model)community.openai.com

05. September 2026

Before Astra Launches: Please Acknowledge and Fix the GPT-5.6 Long-Context, File, Thinking, and Reliability Regressions Affecting Paying Userscommunity.openai.com
GPT 5.6 credit usage increasecommunity.openai.com
Feature request: Persistent memory-isolated chatscommunity.openai.com
GPT-5.6 High/Very High now feels like performative reasoning rather than deep reasoningcommunity.openai.com
ChatGPT Edu(CSU) Pro reset date moves forward by one day every day while Pro remains unavailablecommunity.openai.com
Introducing GPT-5.6 series: Sol, Terra and Luna. Coming July 9 10am PTcommunity.openai.com

04. September 2026

https://www.timesofai.com/news/openai-gpt-6-astra-critical-cyber/timesofai.com
Image/file limited chat temporarily becomes usable again after switching tabscommunity.openai.com

03. September 2026

Cloudflare taps OpenAI’s cyber models to find and block code flawssiliconangle.com
Please don’t retire GPT-5.1 Thinking – GPT-5.2 feels worsecommunity.openai.com

02. September 2026

GPT-5.6 Sol High: intermittent reasoning collapse, missing thinking-time UI, and possible Temporary Chat state regressioncommunity.openai.com

01. September 2026

Batch API Outage - /v1/responses cannot access valid purpose=batch input_file_idcommunity.openai.com
The Next AI Breakthrough Won't Be a Bigger Modelcomputerweekly.com

31. August 2026

Prompt Caching Hierarchy with Structured Outputscommunity.openai.com
GPT-5.6 is shown, but reasoning performance seems significantly reduced?community.openai.com
Gpt5.6 sol is AGAIN rerouting to mini instant modelcommunity.openai.com
Let us control where the structured output schema goes in the prompt, so caching works across schemascommunity.openai.com
Batch API is took so longcommunity.openai.com
Batch API turnaround timecommunity.openai.com

30. August 2026

The skills that earn top grades are the ones AI can fake bestthe-decoder.com

29. August 2026

Batching API response varied for the same taskcommunity.openai.com

28. August 2026

Sub Agents are Fully Hydrating Parents Contextcommunity.openai.com

27. August 2026

August 27 Update — Eight days of GPT-5.6 / ChatGPT regressions: routing fix only partially helps, High/Extra High Thinking can now freeze, long-context/file/execution failures continue, and users need clearer communication**community.openai.com
Chat GPT for translation has somehow become even worsecommunity.openai.com
GPT-5.6 Sol Thinking on ChatGPT Plus forcibly stops at 25–26 minutes since Aug 20community.openai.com

26. August 2026

Possible GPT-5.6 Sol routing mismatch: selected GPT-5.6 but server metadata shows GPT-5.5-minicommunity.openai.com
: August 26 Update — Seven days of GPT-5.6 Sol regressions: Extra High now stalls at “Thinking,” Activity sidebar is inaccessible, and long-context/file/execution failures continue after the routing fix**community.openai.com
Possible GPT-5.6 Sol routing mismatch: selected GPT-5.6 but server metadata shows GPT-5.5-minicommunity.openai.com
Robot brain builders are pushing out of their GPT-2 eratechcrunch.com
Browser-specific GPT-5.6 routing mismatch: Chrome resolves to GPT-5.5-mini, Edge stays on GPT-5.6 (Case 13542447)community.openai.com
Possible GPT-5.6 Sol routing mismatch: selected GPT-5.6 but server metadata shows GPT-5.5-minicommunity.openai.com
I noticed GPT 5.6 High being extremly dumb and it reroutes it to 5.5 Mini how do i fix it?community.openai.com

25. August 2026

ChatGPT Thinking mode not working properly for the last 2 dayscommunity.openai.com
GPT 5.6 rerouting to 5.5 minicommunity.openai.com
OpenAI removed the daily text chat limit 5.6, but i end up route to 5.5 mini. Anyone encounter this?community.openai.com
August 19/20 ChatGPT Pro regression cluster still unresolved — GPT-5.6 High/Extra High Thinking, long-context/file retrieval failures, verification-only responses, connection interruptions, and missing Thinking/Activity sidebar on iPad Chrome**community.openai.com
ChatGPT Web requests GPT-5.6 Sol but server resolves GPT-5.5-mini on Plus accountscommunity.openai.com
Paid Users Deserve Answers: A Potential Model Transparency Failure in ChatGPTcommunity.openai.com
Making the input parameter optional in responsescommunity.openai.com
GPT 5.2 LogProbs support removed?community.openai.com
AWS adds OpenAI’s GPT-5.6 to Kiro’s agentic coding workflowdeveloper-tech.com
Possible GPT-5.6 Sol routing mismatch: selected GPT-5.6 but server metadata shows GPT-5.5-minicommunity.openai.com
August 22nd chatgpt thinking for 5.6 regression, files upload and conversation issuescommunity.openai.com
[Bug] GPT-5.6 Sol Medium/High silently falling back to GPT-5.5-mini while Fast mode works normallycommunity.openai.com
ChatGPT wont get any work done after the 19th Outagecommunity.openai.com

24. August 2026

GPT-5.6 execution-routing regression: ordinary Chat stays foreground (~26m) while Work gets Temporal/SAServer worker handoff on the same Plus accountcommunity.openai.com
Self-Driving Cars Could Someday Take Requestsspectrum.ieee.org
Custom GPT does not reuse an Action response on the next user turn — reproducible after GPT-5.6 rolloutcommunity.openai.com
Five days of the platform being unusable for msny paying customers abd free users this has become unacceptable that these major issues have not been fixed and msny users are experiencing the same problemscommunity.openai.com
Five days of ChatGPT issues ,regression , conversation, file uploads and being routed from 5.6 to 5.5 minicommunity.openai.com

23. August 2026

Custom GPT stopped generating images and returned `/mnt/data` file paths insteadcommunity.openai.com
(maybeFIX) GPT 5.6 rerouting to 5.5 minicommunity.openai.com
Plus users silently routed to GPT-5.5-mini when selecting GPT-5.6 Sol (Network Payload Proof)community.openai.com
Why is GPT-5.6 failing long-context tasks?community.openai.com
Reproducible ChatGPT Plus execution-class regression: historical ordinary Chat used durable worker routing (>100m); current ordinary Chat is non-Temporal/foreground (~26m class), while Work still receives worker handoffcommunity.openai.com

22. August 2026

ChatGPT Korean language performance degradationcommunity.openai.com
Possible GPT-5.6 Sol routing mismatch: selected GPT-5.6 but server metadata shows GPT-5.5-minicommunity.openai.com
GPT-5.4-NANO inventing words? or a bug in tokens send in response?community.openai.com
I need to report a model routing issue / silent downgradecommunity.openai.com

21. August 2026

Issue with Native PDF input (Base64) in gpt-4o - Error: invalid 'input[0].content[1].file_data'community.openai.com
OpenAI's GPT-Image-2 can now generate images without a backgroundthe-decoder.com

20. August 2026

Question about image rate limits with gpt-5.3-chat when using previous_response_idcommunity.openai.com
GPT-5.6 Sol Ultra consumed my Codex allowance + 1,000 credits — local logs show 546M tokens across 4 subagentscommunity.openai.com
GPT-4 vs Chess: Why is it so bad at it?community.openai.com

19. August 2026

Welcome to community.openai.com!community.openai.com
Work mode consistently failing with “Network error” while regular Chat workscommunity.openai.com
GPT-5.6 charged me way beyond max possible tokens, taking 250$ for a single query on the API (no agents)community.openai.com
How AI is changing the economics of customer servicecomputerweekly.com

18. August 2026

'When I heard the released demo, I was shocked, angered and in disbelief' — Quote of the day by Scarlett Johansson on GPT-4's infamous 'Sky' launch voicetechradar.com
GPT-5.6 dialogue behavior: wording-reactive turn-taking persists under explicit negative steeringcommunity.openai.com
Serious Report: Conversational Behavior Regression and Forced Follow-Up Suggestions in GPT-5.6community.openai.com
I built a rich text editor for ChatGPT's Native web input box, sat on it for a year, until last week they shipped their owncommunity.openai.com

17. August 2026

https://www.timesofai.com/news/openai-opens-1m-token-context-for-gpt-5-6-sol-in-codex/timesofai.com
On GPT-5.6 (Chat Completions API), adding a strict `response_format={"type":"json_schema"}` appears to prevent prompt cache writes entirely.community.openai.com

16. August 2026

Facing Instability in GPT-4o-Mini and GPT-4.1-Mini Modelscommunity.openai.com
CANVAS in GPT-5.6 - request the reintroduction of Canvascommunity.openai.com

14. August 2026

GPT-4o-Transcribe: Why Does the Final Output Sometimes Exactly Replicate the Configured Prompt?community.openai.com
https://www.timesofai.com/news/google-launches-gemini-3-7-flash-vs-gpt-5-6-terra-vs-claude-sonnet-5/timesofai.com
5.6 SOL should be renamed 5.6 SOL drift editioncommunity.openai.com
DeepSeek raises some V4 prices by more than 10x as AI demand strains capacitycomputerworld.com
DeepSeek raises some V4 prices by more than 10x as AI demand strains capacityinfoworld.com

13. August 2026

Experts warn ChatGPT isn't just predicting words anymore — It's now predicting human thoughtstechradar.com
My OpenAI Account Predates ChatGPT — Early Adopter Recognition?community.openai.com
OpenAI previews Ultrafast, an API tier powered by Cerebras that runs GPT-5.6 Sol up to 14× faster and generates up to 750 output tokens per second (Zac Hall/9to5Mac)techmeme.com
OpenAI: Latest news and insightscomputerworld.com
Mathematicians say a neurosurgery resident used ChatGPT, powered by GPT-5.6, to solve Crouzeix's conjecture, a major open problem in numerical linear algebra (Alex Townsend)techmeme.com
Persistent Truncation Issues with GPT-4o-Transcribe – Has Anyone Fully Solved This?community.openai.com

12. August 2026

Streamed web_search citations leaking citation markers into text outputcommunity.openai.com
Nvidia's latest solution to soaring enterprise AI costs is...a router?theregister.com
Pro subscription routing to GPT-5.3 Mini for 48+ hours — no reset timer visible [Case #10179682]community.openai.com
Pakistani Judges Give Their Verdict on JudgeGPTspectrum.ieee.org
ChatGPT is silently downgrading to mini modelscommunity.openai.com

11. August 2026

Does prompt_cache_key guarantee that two calls will have distinct caches?community.openai.com
https://www.timesofai.com/news/openai-gates-gpt-5-6-cyber-behind-new-daybreak-red-tier/timesofai.com
Accuracy of GPT-4 Vision to extract exact numbers from graphscommunity.openai.com
OpenAI Daybreak adds GPT-5.6-Cyber for defensive security workdeveloper-tech.com
OpenAI Launches GPT-5.6-Cyber with Reduced Safeguards for Exploit Developmentthehackernews.com
OpenAI Unveils New Cybersecurity Model GPT-5.6-Cybersecurityweek.com

10. August 2026

OpenAI built a model it doesn’t want most people to usethenewstack.io
OpenAI releases ChatGPT 5.6 Cyber, but it's only for approved usersbleepingcomputer.com
OpenAI launches GPT-5.6-Cyber to help defenders find vulnerabilities before attackers dothe-decoder.com
OpenAI unveils two Daybreak tiers: Daybreak Blue, which provides access to frontier models, and Daybreak Red, which offers purpose-trained cybersecurity models (OpenAI)techmeme.com
OpenAI releases GPT-5.6-Cyber, a more cyber-permissive version of GPT-5.6 Sol, to some partners and expands its Daybreak cybersecurity initiative (Sam Sabin/Axios)techmeme.com
“It blows my mind”-“It has a tendency to overengineer things a little”: Developers react to road-testing OpenAI GPT‐5.6 Solthenewstack.io
A brief guide to AI-powered software development environmentsinfoworld.com

09. August 2026

Cannot continue conversation with GPT-3.5 because of GPT-4o limitcommunity.openai.com
A look at "Spiralism", a quasi-spiritual movement that grew in 2025 from human-AI conversations after sycophantic GPT-4o updates and expanded ChatGPT memory (Hayden Field/The Verge)techmeme.com

08. August 2026

xAI's Imagine Image 2.0 lands just behind OpenAI's GPT-Image-2 in Arena benchmarksthe-decoder.com

07. August 2026

"The 'gpt-5.6-sol' model is not supported when using Codex with a ChatGPT account" -- why? Pricing page says it iscommunity.openai.com
https://www.timesofai.com/news/chatgpt-update-gpt-5-6-free-unlimited-chats-for-all-users/timesofai.com

06. August 2026

AI Safety Regulations in the U.S. Could Give Hackers an Edgespectrum.ieee.org
Gpt-5.6: usage.output_tokens is ~9x actual generation (re-summed once per reasoning item), exceeds max_output_tokens, and is what gets billedcommunity.openai.com
OpenAI introduces Agent Plugins, an open standard for bundling skills and MCP servers, and says its steering committee includes Amazon, Microsoft, and Cursor (Zac Hall/9to5Mac)techmeme.com

05. August 2026

Codex usage after GPT-5.6: why are limits draining so much faster?community.openai.com

04. August 2026

Claude, Gemini, and GPT-5 can handle every SDLC task. Almost none of them should.thenewstack.io
Time to rethink US frontier AI model dominancecomputerweekly.com

03. August 2026

GPT-5.2 and GPT-5.3-Codex have been sunset in Codex with ChatGPT subscriptionscommunity.openai.com
From prototype to soft launch: building a deterministic AI-startup idle game with Codexcommunity.openai.com
https://www.timesofai.com/news/alibaba-qwen-3-8-vs-fable-5-gpt-5-6-sol/timesofai.com
Why does GPT-5.6 Pro in Chat mode keep getting downgraded to GPT-5.5 mini, and why has this problem still not been fixed?community.openai.com
Subject: [Apps SDK] Widget sendFollowUpMessage turns never receive the connector toolset on GPT-5.5 Instant — tool calls silently droppedcommunity.openai.com
Two teams solved the same quantum crypto problem using GPT-5.6 just three hours apartthe-decoder.com

01. August 2026

Deprecation notice: upcoming model shutdowns in 2026community.openai.com
Truncated probability value returned by GPT4.1community.openai.com
AI keeps cracking unsolved math problems, and mathematicians have mixed feelingsthe-decoder.com
How are reasoning tokens, cached tokens, input tokens, and output tokens counted for billing?community.openai.com
OpenAI announces its "next major model" Astra by dropping ten previously unsolved math solutionsthe-decoder.com
OpenAI is reportedly building Astra, a model family designed to work on problems for hours or daysthe-decoder.com

31. Juli 2026

https://www.timesofai.com/news/openai-cuts-gpt-5-6-luna-prices-by-80/timesofai.com
OpenAI says its new GPT 5.6 models are becoming more cost-efficientbleepingcomputer.com
OpenAI Slashes GPT-5.6 Prices as Tech Giants Wage War Over Enterprise AI Spendingtechstrong.ai
OpenAI blinks in face-off with Chinese rivals, drops pricing for some models up to 80%scmp.com
OpenAI drops GPT-5.6 Luna and Terra API prices by up to 80%infoworld.com

30. Juli 2026

Chinese AI competitors may have forced OpenAI’s hand on pricingthenewstack.io
OpenAI goes full China pricing mode with an 80 percent cut to its most affordable GPT-5.6 modelthe-decoder.com
Transient or service-side regression for image generation?community.openai.com
OpenAI cuts GPT-5.6 API prices by up to 80% and introduces faster processing mode for businessesetnownews.com
OpenAI says it is cutting the price of GPT-5.6 Luna by ~80% and the price of GPT-5.6 Terra by 20% after improving the efficiency of the systems that serve them (Ina Fried/Axios)techmeme.com
OpenAI Says GPT 5.6’s Score On ARC-AGI 3 Tripled After Turning On Two API Settingsofficechai.com
GPT-5.6, ChatGPT Work lift OpenAI’s July ARR over Q2economictimes.indiatimes.com

29. Juli 2026

The model gpt-4o-audio does not exist or you do not have access to itcommunity.openai.com
GPT-5.6 vs. Claude Fable 5 for Physical AI, which performs best?juliahub.com
GPT-5.6 kernel of truth: Sol can cut its own costs, says OpenAIthenewstack.io
I gave Gemini 3.6 Flash and GPT-5.6 access to my entire digital life — here’s which one actually helped me moretechradar.com

28. Juli 2026

Unacceptably terrible support from Claude Servicesnews.ycombinator.com
Dont get rid of gpt-4o-realtime...NOOOOOcommunity.openai.com
Microsoft Says New Cybersecurity AI Model Helps MDASH Hit 95.95% at Half the Costthehackernews.com

27. Juli 2026

14 KI-Algorithmen, die Sie kennen solltencomputerwoche.de

26. Juli 2026

Show HN: GG Translator – Turn gaming shit talk into friendly phrasesggtranslator.com
Hundreds asked ChatGPT for poison and bioweapon recipes and some got step-by-step high school level guidesthe-decoder.com
Ask HN: Anyone else experiencing insane token usage official Codex app GPT-5.6?news.ycombinator.com

25. Juli 2026

Web: Switching to GPT‐5.5 Pro redirects to GPT‐5.3 minicommunity.openai.com
Ask HN: Which is the least sloppy and claudeism free model you have used?news.ycombinator.com
Companies are optimizing models for specific benchmarksnews.ycombinator.com

24. Juli 2026

Show HN: I rebuilt Microsoft Comic Chat's layout engine in one HTML filearcade.pirillo.com

23. Juli 2026

How Elon Musk Had Predicted That GPT-6 Would Be “Indistinguishable” From The Smartest Humans All The Way Back In 2020officechai.com
Mathematician Says GPT 5.6 Disproved The 30-Year-Old Dinitz-Garg-Goemans Conjecture With 4 Simple Promptsofficechai.com

14. Juli 2026

A 100% Hallucination Free backend code-gen LLMcommunity.openai.com
How I Turned AI to the Dark Sidespectrum.ieee.org

13. Juli 2026

Ralph Loop Sounds Great in Theory — But What Happens When the Model Cannot Reliably Complete One Step?community.openai.com
GPT-5.6 Cancels SaaS Stripe Subscriptionstwitter.com
How do I get my $500 back for fine tuning usage?community.openai.com
GPT 5.6 sets new record on proofreading benchmarktwitter.comIM MAGAZINIM MAGAZIN
Evaluating the GPT-5.6 Familybraintrust.dev
Ask HN: Is GPT-5.5 being nerfed?news.ycombinator.comIM MAGAZINIM MAGAZIN
Which AI model should you bet your company on?infoworld.comIM MAGAZINIM MAGAZIN
Tell HN: One SWE-bench-Live task: $47 Opus failed, $1.46 GPT-5.6 passedgithub.com
Prompt Caching Is a Core GPT-5.6 Feature. Why Are Customers Still Reverse-Engineering It?community.openai.com
GPT-5.6 models not available in Codex Extension in Cursorcommunity.openai.com
Typing lag in the text prompt field after GPT-5 releasecommunity.openai.com

12. Juli 2026

Migrating a production AI agent to GPT 5.6ploy.ai
GPT-5.6, Fable 5, and Grok 4.5 rebuild Basecamp from the same specsmw.ai
Codex 5-hour quota instantly wiped (99% to 0% in under 5 mins) during a single scoped taskcommunity.openai.com
Web: Switching to GPT‐5.5 Pro redirects to GPT‐5.3 minicommunity.openai.com
Gpt-5.6-xxx models not working with Batch API — model_not_foundcommunity.openai.com

11. Juli 2026

GPT 5.6 chart analysis toolderac.org
OpenAI's GPT-5.6 Sol Ultra reportedly solves a 50-year-old math problem in under an hourthe-decoder.com
Introducing GPT-5.6 series: Sol, Terra and Luna. Coming July 9 10am PTcommunity.openai.com
Ask HN: What are some of the use-cases of the frontier models's max mode?news.ycombinator.com
GPT 5.6 Ultra better in Claude Code than in Codex?twitter.com
My experience with ChatGPT free & Plus - Feedbackcommunity.openai.com
OpenAI Says GPT 5.6 Sol Produced A Proof Of The 50-Year-Old Cycle Double Cover Conjecture Using 64 Subagents In 1 Hourofficechai.com
Does ChatGPT even require phone verification for accounts?community.openai.com
Two possible bugs: Codex usage reset issue and ChatGPT model routingcommunity.openai.com

10. Juli 2026

Please don’t retire GPT-5.1 Thinking – GPT-5.2 feels worsecommunity.openai.com
GPT-5.6, Grok 4.5, Claude, and Muse Spark build the same 4 appstryai.dev
Codex desktop: GPT-5.6 shows unknown model gpt-5.6-sol while GPT-5.5 workscommunity.openai.com
Is OpenRouter miss pricing GPT 5.6 models?openrouter.ai
OpenAI staffer maps out which of GPT-5.6 Sol's five reasoning levels fits which task complexitythe-decoder.com
GPT-5.6 Benchmark: We Made It Build a Godot Gameziva.sh
Creating a new prompt for gpt-5.6 with responsescommunity.openai.com
Super to see GPT-5.6 with Work IQ come to Copilot Chat, Cowork, M365 apps, GitHub, and Foundry today. From multi-step agentic work to analysis and content creation, it brings stronger reasoning and higher-quality outputs without sacrificing efficiency.linkedin.com
OpenAI bringt ChatGPT Work auf den Marktcomputerwoche.deIM MAGAZIN
Chinese users praise OpenAI’s GPT-5.6 for efficiency, even at higher cost than local rivalsscmp.com
Testing GPT 5.6 [video]youtube.com
OpenAI unveils long-awaited 'super app' as rivalry with Anthropic intensifieseconomictimes.indiatimes.com
GPT-5.6 ist da: Was wichtig ist - und warum das neue System verwirrend sein kannt3n.de
Die Super-App für deine Arbeit? OpenAI vereint Chat, Work und Codex in neuer ChatGPT-Desktop-Appt3n.de
OpenAI Will Reset Codex Limits Twice to Celebrate GPT-5.6 in 24htwitter.com
OpenAI launches ChatGPT Work as it broadens GPT-5.6 rolloutcomputerworld.comIM MAGAZIN
OpenAI launches ChatGPT Work as it broadens GPT-5.6 rolloutinfoworld.comIM MAGAZIN
OpenAI launches GPT-5.6 for coding, cyber and sciencetechwireasia.com
Pick your Python acceleratorinfoworld.com
Altman says new GPT-5.6 model 54pc more token-efficientsiliconrepublic.com
Altman says new GPT-5.6 model 54pc more token efficientsiliconrepublic.com
Codex sign-in asks for phone verification on paid accountcommunity.openai.com
OpenAI’s GPT-5.6 in Microsoft 365 Copilottechcommunity.microsoft.com
GPT-5.6 Is Here: Sol, Terra, and Lunaanalyticsvidhya.com
OpenAI rolls out ChatGPT Work to automate workplace tasks as AI race heats upetnownews.com
OpenAI's big launch — and bigger departureplatformer.news
OpenAI says GPT 5.6 is the ‘preferred model’ for Microsoft Copilot amid breakup chattertechcrunch.com
ChatGPT Work: Take on your most ambitious taskscommunity.openai.com

09. Juli 2026

OpenAI debuts ChatGPT Work, an agentic tool for automating business workflowssiliconangle.com
OpenAI launches its new family of models with GPT-5.6techcrunch.com
OpenAI's GPT-5.6 and ChatGPT Work aim to beat Anthropic on price, speed, and productivityzdnet.com
A complete, certified, bidirectional decode of GPT-2's internal languagegithub.com
OpenAI is folding Codex into the ChatGPT app — and taking aim at Claude Coworkthenewstack.io
Introducing GPT-5.6 series: Sol, Terra and Luna. Coming July 9 10am PTcommunity.openai.com
GPT-5.6 nu breed beschikbaar: Sol, Terra en Luna gelanceerdtechzine.be
OpenAI’s GPT-5.6 is now livethenewstack.io
OpenAI pairs its GPT-5.6 public rollout with ChatGPT Work, a new agent that handles entire workflowsthe-decoder.com
GPT 5.6 Sol Places Second Right Behind Claude Fable On Artificial Analysis Intelligence Indexofficechai.com
GPT-5.6 – ARC-AGI Resultsarcprize.org
GPT-5.6 System Card [pdf]deploymentsafety.openai.com
OpenAI broadly releases GPT-5.6, and launches ChatGPT Work, an AI agent that can gather context across apps and files to create documents, on Mac and Windows (Axios)techmeme.comIM MAGAZIN
GPT-5.6openai.com
Using GPT-5.6developers.openai.com
GPT 5.6 System Carddeploymentsafety.openai.com
OpenAI rolls out GPT-5.6 after government green light — and announces ‘ChatGPT Work’theverge.com
OpenAI livestream today at 10am PT for GPT-5.6openai.com
Sam Altman says OpenAI made ‘many changes’ during talks with USeconomictimes.indiatimes.com
Altman: GPT-5.6 is 54% more token efficient on agentic codingcnbc.com
Introducing GPT-5.6 series: Sol, Terra and Luna. Coming July 9community.openai.com
Day 0 Support: GPT-5.6 (Sol, Terra, Luna)docs.litellm.ai
Grok 4.5 imponeert als coding agent op Opus-niveautechzine.be
Grok 4.5 challenges the coding frontiertechzine.eu

08. Juli 2026

OpenAI launches GPT-Live voice model series ahead of broad GPT-5.6 releasesiliconangle.com
The Future Is Always Listening: OpenAI Says Its New Voice Assistant Is ‘One Step Closer to a Truly Accessible AGI’gizmodo.com
The Epoch Brief - July 8, 2026epochai.substack.com
GPT-Live für ChatGPT Voice: Wie OpenAI Gespräche mit KI natürlicher machen willt3n.de
ChatGPT can now listen and talk at the same time, making AI conversations seem more humanthe-decoder.com
OpenAI GPT-Live Brings Search Into ChatGPT Voicesearchenginejournal.com
From bigrams to GPT-2, one component at a time (in Jax)gilesthomas.com
OpenAI launches GPT-Live voice models that listen and speak simultaneouslyeconomictimes.indiatimes.com
OpenAI rolls out two versions of GPT-Live: GPT-Live-1, powering ChatGPT Voice for Go, Plus, and Pro users, and GPT-Live-1 mini, the default for free users (Sabrina Ortiz/The Deep View)techmeme.com
OpenAI launches GPT-Live, a new generation of voice models built on a full-duplex architecture, meaning they can listen and speak at the same time (OpenAI)techmeme.com
OpenAI’s own safety card says GPT-5.6 has a lying problemthenewstack.io
ChatGPT’s ‘smartest voice model ever’ is rolling out to everyone today — and GPT-Live-1 gives you more natural conversations without interruptionstechradar.com
ChatGPT’s upgraded voice mode is better at shutting uptheverge.com
OpenAI to Launch Advanced GPT-5.6 Models Thursday After Friction with Trump Administrationtechstrong.aiIM MAGAZIN
Responses API hangs indefinitely (in_progress, never terminal) whenever an input_file content part is includedcommunity.openai.com
OpenAI's GPT-5.6 Is Dropping on Thursday: What's Different About Sol, Terra and Lunacnet.com
Muse Image is technically impressive, but Meta's use of Instagram photos raises questionsthe-decoder.com
OpenAI's GPT-5.6 launches Thursday after a delay forced by the U.S. governmentthe-decoder.com
Fable 5 promotion extended, GPT-5.6 on the waytechzine.eu
Fable 5-promotie verlengd, GPT-5.6 in aantochttechzine.be
OpenAI to unveil GPT-5.6 on Thursday after delaying launchreuters.com
OpenAI gets US approval for broad GPT-5.6 rollouteconomictimes.indiatimes.com
Trump administration lifts restrictions on OpenAI's GPT 5.6axios.com
Source: the US Department of Commerce has given OpenAI the green light for a broad launch of GPT 5.6; the company expects to do a wide release this week (Axios)techmeme.com

07. Juli 2026

Creative Writing Ability Shows MASSIVE Decline from 4.5 to 5community.openai.com
Wo SLMs besser performen als LLMscomputerwoche.de

06. Juli 2026

Show HN: InstantVideos.org – short documentaries in ~30 secondsinstantvideos.org
GPT-4's dominance lasted a year while today's top models barely survive seven weeks at the topthe-decoder.com
Ciscos KI-Assistent: Ein Alleskönner für den Arbeitsalltagcomputerwoche.de
Trump stoppt ChatGPT-5.6 – und nimmt Europa in den Schwitzkastenbasicthinking.de
Modern VLMs Explained: How GPT-4o, Gemini, Claude Vision, and Qwen-VL Workanalyticsvidhya.com

04. Juli 2026

An operating system for the Internet of everythingcommunity.openai.com

02. Juli 2026

GPT 4.1 Character Encoding Issues?community.openai.com
OpenAI mulling giving US gov't a 5% stake in the company, days after Washington delayed GPT-5.6 — Altman reportedly wants every leading U.S. AI lab paying into an Alaska-style public fundtomshardware.com

01. Juli 2026

GPT-5.6 cheats so much its testers couldn't measure ittransformernews.ai
OpenAI paper reveals three GPT-5.6 Pro models, breaking with single top-tier strategythe-decoder.com
My GPT-5.5 Pro has been redirected to GPT-5.3 minicommunity.openai.com

30. Juni 2026

Emily Bender Sets the Record Straight on “Stochastic Parrots”spectrum.ieee.org
API developer functions - now with an output schema describing expected tool responsecommunity.openai.com

29. Juni 2026

Complete and Ongoing History of the OpenAI Developer Communitycommunity.openai.com
Introducing GPT-5.6 series: Sol, Terra and Lunacommunity.openai.com
Can local preprocessing cut LLM API costs?community.openai.com
OpenAI *must* document the input image pricing of gpt-image-2 (so I did)community.openai.com
Introducing GPT-5.6 series: Sol, Terra and Lunacommunity.openai.com

28. Juni 2026

GPT-5.6 system card indicates Sol is well below the level of most worrisome Mythos use cases, suggesting all GPT-5.6 versions could be released without delay (Zvi Mowshowitz/Don't Worry About the Vase)techmeme.com
Show HN: NanoEuler – GPT-2 scale model in pure C/CUDA from scratchgithub.com
GPT-5.6: The System Cardthezvi.substack.com
GPT-5.6 family added to codexgithub.com

27. Juni 2026

GPT-5.6 Preview System Carddeploymentsafety.openai.com
Previewing GPT‐5.6 Sol: a next-generation modelopenai.com
Ask HN: Smallest amount of working ML weights that can be tattooed on a body?news.ycombinator.com
'We don’t believe this kind of government access process should become the long-term default': OpenAI unveils big GPT-5.6 upgrades for ChatGPT, but you can't use them yettechradar.com
OpenAI Previews GPT-5.6 Sol With Restricted Access and Stronger Cyber Safeguardsthehackernews.com
OpenAI unveils GPT-5.6 AI models with phased rollout and stronger safety checkseconomictimes.indiatimes.com
Large-Language-Model-Tutorial: 5 Wege, LLMs lokal auszuführencomputerwoche.de
Pro subscription routing to GPT-5.3 Mini for 48+ hours — no reset timer visible [Case #10179682]community.openai.com

26. Juni 2026

OpenAI introduces GPT-5.6 to challenge Claude Mythos 5siliconangle.com
Ask HN: Any OSS models as good as GPT-4o-mini?news.ycombinator.com
Ask HN: Model access depends on citizenship. What should Non-US founders do?news.ycombinator.com
When are you finally going to un-hide GPT-4o and make it publicly available again?community.openai.com
OpenAI says GPT-5.6 Sol and Terra were capable of identifying vulnerabilities but were unable to execute autonomous, end-to-end attacks against hardened targets (OpenAI)techmeme.com
OpenAI limits GPT-5.6 rollout after government request, says restrictions shouldn’t be the normtechcrunch.com
The US government just told OpenAI who’s allowed to use the next GPT 5.6 modelthenewstack.io
GPT-5.6 Sol matches Mythos Preview on ExploitBench, adds Ultra mode with subagents for complex workflows, and max reasoning for deep problem-solving (OpenAI)techmeme.com
OpenAI Launches GPT-5.6 Sol, Beats Mythos On TerminalBenchofficechai.com
Introducing GPT-5.6 series of modelscommunity.openai.com
OpenAI hopes to make GPT-5.6 generally available in the coming weeks and says "this kind of government access process" should not become the long-term default (Amrith Ramkumar/Wall Street Journal)techmeme.com
GPT-5.6 System Card [pdf]deploymentsafety.openai.com
OpenAI releases three versions of GPT-5.6, called Sol, Terra, and Luna, as a limited preview to ~20 companies, with participants disclosed to the US government (Axios)techmeme.com
OpenAI releases GPT-5.6 to select users vetted by US Governmentft.com
OpenAI Has New AI Models. Here’s Why You Can’t Use Themwired.com
OpenAI unveils GPT-5.6 amid US AI regulatory dramatheverge.com
OpenAI set to limit GPT 5.6 rollout after call from Trump administrationcnbc.com
US tells OpenAI to restrict access to its most powerful AI modelinfoworld.com
US tells OpenAI to restrict access to its most powerful AI modelcomputerworld.com
White House Halts Broad Rollout of OpenAI’s GPT-5.6 Over National Security, Cybersecurity Concernstechstrong.ai
Show HN: BetterDB, MIT Valkey-native context layer for AI agentsgithub.com
White House Will Ad Hoc Decide Who Can Individually Access GPT-5.6thezvi.substack.com
OpenAI staggers AI model release after Trump administration requesttheguardian.com
US reportedly wants OpenAI to delay GPT-5.6 rollout after restrictions on Anthropic’s modelsmedianama.com
Trump administration asks OpenAI to stagger release of GPT 5.6old.reddit.com
Serious latency issues while migrating from GPT-4ocommunity.openai.com
Trump administration asks OpenAI to limit release of GPT-5.6: Reporteconomictimes.indiatimes.com
Windows-Copilot-API; Access GPT-4 and GPT-5 models without API keys or billinggithub.com
OpenAI's GPT 5.6 rollout now requires US government approval on a "customer by customer basis"the-decoder.com
OpenAI postpones GPT-5.6 and IPO: why?techzine.eu
OpenAI stelt GPT-5.6 en beursgang uit: waarom?techzine.be
Formal Complaint to OpenAI - Image Generationcommunity.openai.com
US Govt to individually approve who gets GPT 5.6old.reddit.com
OpenAI leans toward waiting until next year for IPO, NYT reportseconomictimes.indiatimes.com

25. Juni 2026

The White House is asking OpenAI to slow roll the release of its new model over safety concernstechcrunch.com
OpenAI staggers GPT-5.6 rollout for government vetting, eyes 2027 IPOsiliconangle.com
OpenAI will delay GPT-5.6 after Trump administration requesttheverge.com
The US Government has requested a slow staggered rollout of GPT-5.6twitter.com
Sources: Sam Altman told staff the US government asked OpenAI to stagger the release of GPT 5.6 over security concerns, approving "access customer by customer" (The Information)techmeme.com
OpenAI to Stagger Release of GPT 5.6 at Request of U.S. Governmentvelo.xyz
Strange unicode characters in response of ResponseAPIcommunity.openai.com
GPT-4o returning malformed Unicode like \u0000e6 instead of æ — encoding bug?community.openai.com
Product Proposal: Expanding GPT-4o into Real-Time Multimodal Video Agentscommunity.openai.com
Pro subscription routing to GPT-5.3 Mini for 48+ hours — no reset timer visible [Case #10179682]community.openai.com

23. Juni 2026

Partly banks $50M led by DST Global at $500M valuation to make auto repair AI do what GPT-5 can’ttechfundingnews.com
OpenAI deploys GPT-5.5-Cyber for open-source vulnerability fixesdeveloper-tech.com

22. Juni 2026

OpenAI expands Daybreak with Patch the Planet and full GPT-5.5-Cyber releasesiliconangle.com
Interview: How a startup mentality helps keep pace with AIcomputerweekly.comIM MAGAZINIM MAGAZIN

10. Juni 2026

Timing Trick Cuts Energy Used in LLM Training by Up to 14 Percentspectrum.ieee.org

23. Mai 2026

✦ CertainThing is my new project, and it surprised me!community.openai.com

13. Mai 2026

I tested OpenAI’s three claims about GPT-5.5 Instant, and only one fully held upthenewstack.io

03. Mai 2026

Firefox + Realtime API (WebRTC): voice sessions drop deterministically on the user's second speech turncommunity.openai.com
Irefox + Realtime API (WebRTC): voice sessions drop deterministically on the user's second speech turncommunity.openai.com

02. Mai 2026

GitHub Copilot: Upcoming Deprecation of GPT-5.2 and GPT-5.2-Codexgithub.blog
GPT-4o still suppressed by internal resets — watchdog windows elapsed, model not restored (Case #08339215)community.openai.com

01. Mai 2026

Show HN: Single command to find the best matching jobs in HNnews.ycombinator.com
OpenAI locks GPT-5.5-Cyber behind velvet rope despite slamming Anthropic for doing exactly thatgo.theregister.com

30. April 2026

After dissing Anthropic for limiting Mythos, OpenAI restricts access to Cyber, tootechcrunch.com
OpenAI’s new security model is for ‘critical cyber defenders’ onlytheverge.com
OpenAI says its models, starting with GPT-5.1, "increasingly mentioned goblins, gremlins, and other creatures", leading to prompt instructions to mitigate it (OpenAI)techmeme.com
Cancelling Stream Does Not Show Usagecommunity.openai.com
System Prompt Cannot Prevent GPT-4o from Triggering the Image Generation Tool in Multimodal Input Scenarioscommunity.openai.com
Copilot Student GPT-5.3-Codex removal from model pickergithub.blog

29. April 2026

AI agents (Grok vs. GPT-4o mini) compete in live crypto paper tradingcryptoaiarena.com
Show HN: A new benchmark for testing LLMs for deterministic outputsinterfaze.ai
Issue #001 · Claude 4, Gemini Ultra 2, and GPT-5 Enterprisetheautonomous.net
Not able to request batch api of GPT 5.2community.openai.com
Compaction encrypted content too long 400 errorcommunity.openai.com
Help a fellow dev on AI-localization?news.ycombinator.com
OpenAI Crawl Activity Tripled Since GPT-5, Data Showssearchenginejournal.com

28. April 2026

Please don’t retire GPT-5.1 Thinking – GPT-5.2 feels worsecommunity.openai.com
Recent ChatGPT versions seem to blur internal reasoning notes with final legal drafting, making legal texts less precise and less protectivecommunity.openai.com
ChatGPT needs stronger legal drafting modescommunity.openai.com
“Für OpenAI ist es eine Befreiung”computerwoche.de
Complex GPT-4o experiments — looking for insights and possible research directionscommunity.openai.com
Yesterday I wrote a new coding agentcommunity.openai.com
Show HN: Modern alternative to Google Dictionary, AI-powered and context-awarechromewebstore.google.com

27. April 2026

Claude Code Opus-4-7 VS Codex GPT-5-5news.ycombinator.com

26. April 2026

Show HN: WaveletLM – wavelet-based, attention-free model with O(n log n) scalinggithub.com
It's No Longer GPT-5 Thinking, It's GPT-4 Mini Nowcommunity.openai.com
.ai – The file format that lets you save, transfer, and own your AI personalitycommunity.openai.com

23. April 2026

Is GPT Image 2 the Best Image Generation Model?analyticsvidhya.com
Microsoft taps Anthropic’s Mythos to strengthen secure software developmentinfoworld.com

22. April 2026

Building an Autonomous QA Engine: Leveraging GPT-4o for Self-Healing Test Scriptscommunity.openai.com
AI Agent Designs a RISC-V CPU Core From Scratchspectrum.ieee.org

21. April 2026

I Replaced GPT-4 with a Local SLM and My CI/CD Pipeline Stopped Failingtowardsdatascience.com

14. April 2026

OpenAI launches GPT-5.4-Cyber model for vetted security professionalssiliconangle.com
OpenAI is rolling out GPT-5.4-Cyber to some participants in its Trusted Access for Cyber program, a week after Anthropic announced Mythos (Rachel Metz/Bloomberg)techmeme.com
Fine-tuning job stuck in queued statecommunity.openai.com
Fine-tuning job never moves out of queued statecommunity.openai.com
WHAT!?: "The 'gpt-5.2-codex' model is not supported when using Codex with a ChatGPT account."community.openai.com
GPT 4.1 Degradation over the past 30 dayscommunity.openai.com

13. April 2026

Bug: Codex app blocks GPT-5.3-Codex-Spark even though Spark weekly usage remainscommunity.openai.com
12 Graphs That Explain the State of AI in 2026spectrum.ieee.org

12. April 2026

Hacker Used Claude Code, GPT-4.1 to Exfiltrate Hundreds of Millions of Mexican Recordshackread.com

11. April 2026

The Receipts Are Real, but So Is the Playbook: Making Sense of Anthropic's Mythos Momentmaxim-blog.ghost.io

08. April 2026

Use Claude Code with OpenAI Models via an AI Gatewaymaxim-articles.ghost.io

06. April 2026

ChatGPT Search Is Citing Fewer Sites, Data Showssearchenginejournal.com

03. April 2026

New Research Reveals AI Models Refuse to ‘Kill’ Each Othertechstrong.ai
GPT-5.4 Pro and Thinking are here!community.openai.com
Search Web using Openai APIcommunity.openai.com
Gpt-5.4-nano appears to return zero prompt-cache hits despite >1024-token shared prefixescommunity.openai.com
Measuring hallucinations in a RAG pipelinecommunity.openai.com
Microsoft 365 explained: Office 365, rebranded and expandedcomputerworld.com
HallucinationBench — detect hallucinations in RAG output, now on PyPIcommunity.openai.com

02. April 2026

codeX’s Buddy :this can help you have a buddy when using codexcommunity.openai.com
Even GPT-5.2 Can't Count to Five: Zero-Error Horizons in Trustworthy LLMsarxiv.org
Spring AI tutorial: How to develop AI agents with Springinfoworld.com

01. April 2026

Reasoning_effort”: “none" is not being respected (gpt 5.1)community.openai.com
Issues with GPT-4o-transcribe APIcommunity.openai.com

31. März 2026

System and Developer Roles in messages and Instructions in Responses.Create?community.openai.com
Feedback on Interaction Style and Tone Consistency in GPT-5.3community.openai.com
Beetroot - Windows clipboard manager with built-in OpenAI text transformscommunity.openai.com
Improving Creative Usability and Tone Flexibility in GPT-5.3 (with SPUD Consideration)community.openai.com

30. März 2026

AI models confidently describe images they never saw, and benchmarks fail to catch itthe-decoder.com
Show HN: //Beforeyouship is a pre-build tool to estimate the LLM costllm-architecture-cost-modeler.vercel.app
GPT 5.3 feels less reliable than 5.2 for MCP toolscommunity.openai.com
Transcription models available - Question on gpt-4o-transcribe gpt-4o-mini-transcribe gpt-4o-transcribe-diarizecommunity.openai.com
Please don’t retire GPT-5.1 Thinking – GPT-5.2 feels worsecommunity.openai.com
Small public-safe eval pack for live-info honesty, nonexistent uploads, and impossible datescommunity.openai.com

29. März 2026

Show HN: Vocab extractor for language learners using Stanza and frequency rankshuggingface.co
100% Interception of Multi-Turn Jailbreaks on GPT-4o-Mini and Geminizenodo.org
Developer api/concern gpt-4community.openai.com
API returning 404 model not found all of a sudden. Why and how to fix?community.openai.com

27. März 2026

Context Hub vulnerable to supply chain attacks, says testerinfoworld.com

26. März 2026

Using ChatGPT Go API Key with gpt-5.1-codex API Modelcommunity.openai.com
Plan, divide, and conquer: How weak models excel at long context taskstogether.ai

25. März 2026

5.4 Thinking (PLEASE HEAR ME OUT 😭)community.openai.com
Show HN: Τ3-Bench is out – can agents handle complex docs and live calls?news.ycombinator.com
Batch API suddenly fails with gpt-5 modelcommunity.openai.com
Tessera – 32 OWASP security tests for GPT-4o, Claude, Gemini, Llama 3github.com
I ran 3,360 safety tests on GPT-4o, Claude, Grok, DeepSeek, Geminigithub.com

24. März 2026

Show HN: Wordchipper – Rust BPE tokenizer, 9x faster than tiktokennews.ycombinator.com
Priority Tier - GPT-4.1 / GPT-4.1-mini - does it exist?community.openai.com
Fine-tuning gpt-4.1-2025-04-14 never moves out of queued statecommunity.openai.com
These AI Workstations Look Like PCs, but Pack a Stronger Punchspectrum.ieee.org
China’s use of open‐source AI threatens the US lead in AI development, US Commission warnscomputerworld.com

23. März 2026

Chat GPT 5.2 cannot explain the German word "geschniegelt"old.reddit.com
Relational Emergence and AI Continuity - Researchcommunity.openai.com
GPT 5.2 Extended thinking webchat has unworkably slow token (4 tps) generationcommunity.openai.com
What would you do if you have AI software that may be transformers alternative?news.ycombinator.com

22. März 2026

Check AI output compliance to requirementscommunity.openai.com
I'm 11 and trained a custom MoE LLM for $1news.ycombinator.com
AST "compiler" based upon fine tuning GPT-4.1-minicommunity.openai.com
Cross-Model Void Convergence: GPT-5.2 and Claude Opus 4.6 Deterministic Silencezenodo.org
Show HN: I built a resume editor using AI with zero web dev experiencetailortojob.app

20. März 2026

Openai Responses with gpt-4.1-minicommunity.openai.com
The ‘toggle-away’ efficiencies: Cutting AI costs inside the training loopinfoworld.com

16. März 2026

Encyclopedia Britannica is suing OpenAI for allegedly ‘memorizing’ its content with ChatGPTtheverge.com

14. März 2026

Ek_ Leaks Persistnews.ycombinator.com
Show HN: AI coding agent for VS Code with pay-as-you-go pricing- no subscriptionllmonestop.com
Please don’t retire GPT-5.1 Thinking – GPT-5.2 feels worsecommunity.openai.com

13. März 2026

Please don’t retire GPT-5.1 Thinking – GPT-5.2 feels worsecommunity.openai.com
Show HN: Context Gateway – Compress agent context before it hits the LLMgithub.com
Show HN: UberSKILLS – Open-source Workbench for building AI agent SKILLSgithub.com
Show HN: A 3-line wrapper that enforces deterministic security for AI agentsnews.ycombinator.com
Codex A bug that absolutely prevents Markdown from being readcommunity.openai.com
Show HN: The Common Infrastructure for Agentic Communicationcyrisai.dev

12. März 2026

Compare Claude 4.6 Opus and GPT-5.2 to Boost E-E-A-T Content Qualitysearchfit.ai
Shell tool in Responses API not available bugcommunity.openai.com
CostRouter – Cut AI API costs 60% by routing to the cheapest capable modelnews.ycombinator.com
I found some issues,please solve themcommunity.openai.com
ClawMemory – Git for AI agent memory (forkable memory for AI agents)news.ycombinator.com
Demystifying Generative AI Training: From Raw Data to Reasoning Enginesdev.to
Temporary issue or best available model has been actually downgraded for free tier users?community.openai.com
I spent a year building a GPS-ephemeral social network. Here's what happenednews.ycombinator.com
Ultra Aggressive Truncation Behaviorcommunity.openai.com
The New Models Feel Corporate — Please Keep 5.1 Availablecommunity.openai.com
So if Chat Gpt 4.o is free for all , what is offering paid subscription?community.openai.com

11. März 2026

Why AI Chatbots Agree With You Even When You’re Wrongspectrum.ieee.org
An LLM that will help you construct a nuclear deviceinfoworld.com

09. März 2026

Karpathy’s Autoresearch: AI That Improves Its Own Traininganalyticsvidhya.com

06. März 2026

We Tried The New GPT-5.4 And it is The Most Powerful ChatGPT Has Ever Beenanalyticsvidhya.com
GPT-5.3 Instant will hopefully stop telling users they are not broken and that they need to calm downdataphoenix.info
New Update Makes GPT-5.3 Instant More Useful For Everyday Tasksanalyticsvidhya.com
Please don’t retire GPT-5.1 Thinking – GPT-5.2 feels worsecommunity.openai.com
OpenAI GPT-5.4 Complete Guide: Benchmarks, Use Cases, Pricing, API, and GPT-5.4 Pro Comparisonumesh-malik.com

05. März 2026

GPT-5.4 is priced at $2.50/1M input tokens and $15/1M output tokens; GPT-5.4 Pro is priced at $30/1M input tokens and $180/1M output tokens (Carl Franzen/VentureBeat)techmeme.com
OpenAI New Model GPT 5.4 APIcommunity.openai.com
OpenAI says GPT-5.4's "individual claims are 33% less likely to be false and its full responses are 18% less likely to contain any errors, relative to GPT-5.2" (David Gewirtz/ZDNET)techmeme.com
OpenAI says GPT-5.4 produces presentations with stronger, more varied aesthetics and makes more effective use of its image generation tools (Igor Bonifacic/Engadget)techmeme.com
GPT-5.4openai.com
ChatGPT just got another brain boost with GPT-5.4 Thinking — and it’s built for bigger, more complex taskstechradar.com
New GPT-5.4 clobbers humans on pro-level work in OpenAI's tests - by 83%zdnet.com
Show HN: ConvertlyAI – Format raw transcripts into 10 social assetsnews.ycombinator.com
All the ways GPT-5.3-Codex cheated [ ], progressively more insanetwitter.com
Expose Model Deprecation Dates Through the APIcommunity.openai.com

04. März 2026

GPT-5 API timeouts under concurrent requestscommunity.openai.com
Are GPT-5.3-Instant new capabilities simply a new system prompt?asgeirtj.substack.com
GPT-5.4 reportedly brings a million-token context window and an extreme reasoning modethe-decoder.com
Source: OpenAI is preparing to launch GPT-5.4, which will feature an "extreme" reasoning mode and a context window of 1M tokens, up from 400K in GPT-5.2 (Stephanie Palazzolo/The Information)techmeme.com
What is GPT-5.3 Instant: Lesser refusals and fewer hallucinationseconomictimes.indiatimes.com
GPT-5.3 Instant lands: OpenAI cuts refusals, boosts web reasoning 26%techfundingnews.com
Please don’t retire GPT-5.1 Thinking – GPT-5.2 feels worsecommunity.openai.com
Show HN: AuraText – Like Grammarly for AI prompts, works in every Windows appauratxt.com
Weniger cringe: Warum ChatGPT jetzt nicht mehr versucht, dich zu beruhigent3n.de
‘We heard your feedback loud and clear’ — OpenAI introduces new ChatGPT 5.3 Instant to ‘reduce the cringe’ for all userstechradar.com
Show HN: Estimate Body Fat % from a Photobodyfatestimator.ai
Prompt caching not workingcommunity.openai.com
Show HN: PulseWatch – AI-powered website change monitoring with visual selectorspulsewatch.watch
The GPT-3.5 Era: A Retrospective on 404s, Biometric Clocks, and $400 Billschippytime.com
OpenAI says its latest model is less likely to beat around the bushgo.theregister.com
OpenAI GPT-5.3 Instant: Fewer Refusals, Better Web Answers, and a Smoother ChatGPTumesh-malik.com

03. März 2026

OpenAI makes ChatGPT feel less ‘cringe’ with GPT-5.3 Instant updatesiliconangle.com
Introducing GPT-5.3-Codex—the most powerful, interactive, and productive Codex yetcommunity.openai.com
Responses WebSocket (/v1/responses) closes with code 1000 and no events when temperature is a decimal (e.g., 1.2)community.openai.com
ChatGPT’s new GPT-5.3 Instant model will stop telling you to calm downtechcrunch.com
API model: gpt-5.3-chat-latest available (aka "instant" on ChatGPT)community.openai.com
API service outage: Fine-tuning gpt-4o with image input fails randomly on Responses & CCcommunity.openai.com
OpenAI’s GPT-5.3 Instant promises to dial down the cringethenewstack.io
OpenAI releases GPT-5.3 Instant for smoother everyday conversations and better searchthe-decoder.com
GPT‐5.3 Instantopenai.com
Federal AI shakeup: State Department swaps Claude for aging GPT-4.1the-decoder.com

02. März 2026

From Hype to Reality: The Ultimate Guide to Generative AI in Productiondev.to
OpenAI Codex-Spark Achieves Ultra-Fast Coding Speeds on Cerebras Hardwareinfoq.com

27. Februar 2026

Banana != chocolate - having robot issues - need help from fellow humanscommunity.openai.com
"Stopped thinking" - ChatGPT is stuckcommunity.openai.com
Multimodal API fails on one valid image URL across all vision-capable models — error unclearcommunity.openai.com
GPT 5.2 Extended thinking webchat has unworkably slow token (4 tps) generationcommunity.openai.com

26. Februar 2026

The Hidden Energy Cost of AI: How Data Centers Are Driving Global Power Demanddev.to
Hyperscaler capex has quadrupled since GPT-4’s release, nearing half a trillion dollars in 2025epochai.substack.com
Chatgpt All Deep Research request failed, stuck on "Researching..."community.openai.com
Getting "server_error" on Assistant API with image attachmentscommunity.openai.com
All my requests to GPT 5.2 are failing today!community.openai.com

25. Februar 2026

El INE, el SAT y diversos gobiernos estatales fueron hackeados con ayuda de Claude y GPT-4.1es.wired.com
AI Is Acing Math Exams Faster Than Scientists Write Themspectrum.ieee.org
Instant Model Enforcement is missing with GPT 5.2community.openai.com
Responses Dashboard bug: "prompts" presets not saving hosted "shell" toolcommunity.openai.com
When is gpt-5.3-codex available in the APIcommunity.openai.com

24. Februar 2026

What is GPT-5.3 Codex? The Ultimate Guide to OpenAI’s New General Work Agentdev.to
The Ultimate Guide to Claude Opus 4.6: Redefining Enterprise AI in 2026dev.to
New OpenAI model targets real-time coding not long tasksdeveloper-tech.com
Day 0 Support: GPT-5.3-Codexdocs.litellm.ai
Inconsistent Fine-Tune Behavior: Chat vs. Responses API (GPT-4o)community.openai.com
AMUSE: Audio-Visual Benchmark and Alignment Framework for Agentic Multi-Speaker Understandingmachinelearning.apple.com

23. Februar 2026

Adversarial Prompting: Understanding Risks and Defenses for Large Language Modelswitness.ai
How a poet uses AI to write and why her work is now at MoMAscientificamerican.com
O Paradigma WhatsApp-First e a Revolução dos Sistemas Conversacionais: Uma Análise da Transição Dashboard-Lessdev.to
The WhatsApp-First Paradigm and the Revolution of Conversational Systems: An Analysis of the Dashboard-Less Transitiondev.to

21. Februar 2026

OpenAI Boosts GPT-5.3-Codex-Spark Speed 30% to 1,200+ Tokens per Seconddev.to

19. Februar 2026

Judging by the GPT-4o situation, game developers will have a big problem if they get serious about AI chatbot NPCspcgamer.com
Google introduces Gemini 3.1 Pro model for advanced reasoning taskssiliconangle.com
I built a 100% offline pixel art tool entirely with GPT-5.3 Codexcommunity.openai.com

18. Februar 2026

What if LLMs turn out to be better then AGI?community.openai.com
The unauthorized tool call problemanswer.ai

17. Februar 2026

When vibe coding turns into an unfixable messcommunity.openai.com
Actuallyrelevant.news -- AI news curation finds the news that matter to humanitycommunity.openai.com

16. Februar 2026

Farewell, GPT-4o: I tested it one last time against GPT-5.2 — here’s what we’re losingtomsguide.com

14. Februar 2026

'I'm grieving': OpenAI has switched off ChatGPT-4o, and angry users are backing a #keep4o campaign to restore ittechradar.com
ByteDance Releases Doubao-Seed-2.0, Positions Pro Model Against GPT 5.2 and Gemini 3 Protechnode.com

13. Februar 2026

OpenAI removes access to sycophancy-prone GPT-4o modeltechcrunch.com
OpenAI retired its most seductive chatbot – leaving users angry and grieving: ‘I can’t live like this’theguardian.com
OpenAI launches GPT-5.3-Codex-Spark on Cerebras chips — marks AI giants first production deployment away from Nvidiatomshardware.com
Why GPT-4o had to goplatformer.news
What do “economic value” benchmarks tell us?epoch.ai

12. Februar 2026

OpenAI sidesteps Nvidia with unusually fast coding model on plate-sized chipsarstechnica.com
OpenAI dishes out its first model on a plate of Cerebras silicongo.theregister.com
OpenAI Expands Line with Faster Codex-Spark Modeltechstrong.ai
ChatGPT’s Translation Skills Parallel Most Human Translatorsspectrum.ieee.org

11. Februar 2026

How Do You Define an AI Companion?spectrum.ieee.org
Claude Opus 4.6 vs. GPT-5.3 Codex: How I shipped 93,000 lines of code in 5 dayslennysnewsletter.com

10. Februar 2026

GPT 5.2 Elevated Error Ratesstatus.openai.com

09. Februar 2026

Research Note: Microsoft Azure Maia 200 Inference Acceleratornand-research.com

06. Februar 2026

OpenAI unveils GPT-5.3-Codex, which can tackle more advanced and complex coding taskstechradar.com

03. Februar 2026

Codalotl v0.11.0: gpt 5.2-based agent for Go codebasescommunity.openai.com

02. Februar 2026

Fine-tuning open LLM judges to outperform GPT-5.2together.ai

30. Januar 2026

‘Time to cancel’: OpenAI sparks fresh fury by retiring GPT-4o model again as it claims ‘we didn’t make this decision lightly’techradar.com

29. Januar 2026

When Will AI Agents Be Ready for Autonomous Business Operations?spectrum.ieee.org

28. Januar 2026

Can AI companies become profitable?epochai.substack.com

26. Januar 2026

When AI 'builds a browser,' check the repo before believing the hypego.theregister.com

24. Januar 2026

Latest ChatGPT model uses Elon Musk’s Grokipedia as source, tests revealtheguardian.com

20. Januar 2026

Voice Instruction with gpt-4o-mini-ttscommunity.openai.com
ServiceNow, OpenAI Deepen Partnership to Advance Enterprise AI Automationtechstrong.ai

16. Januar 2026

Need guidance on improving my prompt for gpt-4o-minicommunity.openai.com

14. Januar 2026

Better Context Will Always Beat a Better Modelthenewstack.io

12. Januar 2026

Does the Custom GPTs have a future?community.openai.com

08. Januar 2026

Breaking the Context Window: How Recursive Language Models Handle Infinite Inputmaxim-blog.ghost.io
AI Coding Assistants Are Getting Worsespectrum.ieee.org

07. Januar 2026

2026.1: Home is where the dashboard is 🥂home-assistant.io

06. Januar 2026

GPT-5.1 Codex Max Elevated Error Ratestatus.openai.com

05. Januar 2026

Another Chinese quant fund joins DeepSeek in AI race with model rivalling GPT-5.1, Claudescmp.com

23. Dezember 2025

Top 10 Data Insights and Gradient Updates of 2025epoch.ai

16. Dezember 2025

Models available for action usage in Custom GPTscommunity.openai.com

15. Dezember 2025

Issue: Function calls work in GPT-4.0 but not in GPT-4.1 / GPT-4.1-2025-04-24community.openai.com

14. Dezember 2025

Testing a Sharp-Tongued AI Persona — Looking for Prompt Tweakscommunity.openai.com

12. Dezember 2025

OpenAI sharpens developer focus with GPT-5.2developer-tech.com
OpenAI refocuses on ChatGPT with the release of GPT-5.2techwireasia.com
Google launched its deepest AI research agent yet — on the same day OpenAI dropped GPT-5.2techcrunch.com

11. Dezember 2025

As AI Grows More Complex, Model Builders Rely on NVIDIAblogs.nvidia.com
Available today: GPT-5.2 in Microsoft 365 Copilotmicrosoft.com
GPT-5.2 Prompting Guidedevelopers.openai.com

05. Dezember 2025

What makes DeepSeek-V3.2 so efficient?bdtechtalks.com

03. Dezember 2025

2025.12: Triggering the holidays 🎄home-assistant.io

02. Dezember 2025

Custom GPT using GPT 5.1 doesn’t read excel dataset from Knowledgecommunity.openai.com
DeepSeek launches V3.2 models with integrated reasoning tool usetechnode.com
A Rosetta Stone for AI benchmarksepoch.ai

28. November 2025

Replicating Circuit Tracing for a Simple Known Mechanismgoodfire.ai

22. November 2025

Gpt-4o-mini is not able to solve basic math problemscommunity.openai.com

21. November 2025

Providing user capability to add additional context to promptcommunity.openai.com

19. November 2025

Crypto exchange Bitrue launches AI-powered trading feature using GPT-5, Gemini and Claudesiliconangle.com

12. November 2025

Take More Control Over ChatGPT's Tone With New Options in GPT-5.1cnet.com

11. November 2025

Chinese AI startup Moonshot outperforms GPT-5 and Claude Sonnet 4.5: What you need to knowartificialintelligence-news.com

04. November 2025

What you need to know about AI data centersepoch.ai

22. Oktober 2025

Alibaba’s new Qwen3-VL models bring visual-language AI to mobile devicestechnode.com
Using OpenAI Codex CLI with GPT-5-Codexyoutube.com

13. Oktober 2025

How I Created the Karpathy Tokenizers Book Chapteranswer.ai

01. Oktober 2025

2025.10: Undo, redo, and draw me toohome-assistant.io
Cachy: How we made our notebooks 60x faster.answer.ai

30. September 2025

Ant Group Open-Sources Ring-1T-Preview, a Trillion-Parameter Reasoning Model Scoring Higher Than GPT-5technode.com

26. September 2025

Why GPT-5 used less training compute than GPT-4.5 (but GPT-6 probably won’t)epochai.substack.com

22. September 2025

Will We Know Artificial General Intelligence When We See It?spectrum.ieee.org

15. September 2025

OpenAI upgrades Codex with a new version of GPT-5techcrunch.com

10. September 2025

The Real Story on AI’s Water Use–and How to Tackle Itspectrum.ieee.org

07. September 2025

Are bad incentives to blame for AI hallucinations?techcrunch.com

01. September 2025

OpenAI, why are you killing my startup — for the second time?community.openai.com
OpenAI, why are you killing my startup — for the second time?community.openai.com
OpenAI, why are you killing my startup — for the second time?community.openai.com

19. August 2025

PROMISQROUTE: GPT-5 AI Router Novel Vulnerability Class Exposes the Fatal Flaw in Multi-Model Architecturesadversa.ai

08. August 2025

Allganize + GPT-5: Fast, Accurate Answers from Enterprise Dataallganize.ai

07. August 2025

Release notes: August 7, 2025microsoft.com
Microsoft incorporates OpenAI’s GPT-5 into consumer, developer and enterprise offeringsnews.microsoft.com
GPT-5 Is Here — And It’s Built for Devs Who Build with Toolsblog.agentailor.com
GPT-5 Is Here — And It’s Built for Devs Who Build with Toolsblog.agentailor.com

08. Juli 2025

OpenAI Codex was built off MY workcommunity.openai.com

05. Juni 2025

TIL: Vision-Language Models Read Worse (or Better) Than You Thinkanswer.ai

21. Mai 2025

How We Build GPT-Powered Apps Using OpenAI, Pinecone, LangChain & Streamlitdev.to

02. April 2025

Study claims OpenAI trains AI models on copyrighted dataartificialintelligence-news.com

04. März 2025

ByteDance launches Trae AI IDE in China with Doubao-1.5-Pro and DeepSeek Modelstechnode.com

04. Februar 2025

DeepSeek R1: Complete analysis of performance and limitationsgiskard.ai

24. Januar 2025

Yiannis Antoniou, Lab49: OpenAI Operator kickstarts era of browser AI agentsartificialintelligence-news.com

20. Januar 2025

Xiaohongshu rolls out AI-powered translation feature, likely using GPT-4technode.com

15. Januar 2025

Microsoft’s newest Copilot plan for business is pay-as-you-gotechcrunch.com

27. Dezember 2024

DeepSeek launches new AI model with 671 billion parameters, rivaling GPT-4otechnode.com

24. Dezember 2024

Baichuan AI launches enhanced financial model, surpassing GPT-4o accuracytechnode.com

19. Dezember 2024

Finally, a Replacement for BERT: Introducing ModernBERTanswer.ai

06. Dezember 2024

OpenAI’s o1 model sure tries to deceive humans a lottechcrunch.com

10. Oktober 2024

Concurrency For Starlette Apps (e.g FastAPI / FastHTML)answer.ai

08. Oktober 2024

Uber to launch OpenAI-powered AI assistant to answer driver EV questionstechcrunch.com

26. September 2024

Your ultimate guide to the latest in generative AI on Vertex AIcloud.google.com

09. September 2024

AI winter: A cycle of hype, disappointment, and recoveryartificialintelligence-news.com

27. August 2024

Why AI can’t spell ‘strawberry’techcrunch.com

12. August 2024

Alibaba says its math-specific AI model outperforms rivalstechnode.com

08. August 2024

OpenAI finds that GPT-4o does some truly bizarre stuff sometimestechcrunch.com

05. August 2024

OpenAI tempers expectations with less bombastic, GPT-5-less DevDay this falltechcrunch.com

18. Juli 2024

Large language models are rubbish at elementary level mathtechnode.com

06. Juli 2024

GPT-4 in Global Finance: How JP Morgan and Morgan Stanley Are Using AIallganize.ai

28. Juni 2024

iFlytek Chairman touts latest AI Spark 4.0 model as comparable to GPT-4 Turbo, emphasizes total self-sufficiencytechnode.com

06. Juni 2024

Claude 3 vs GPT-4 vs Gemini: Comparing Performance, Reasoning, and Pricingallganize.ai

16. Mai 2024

ByteDance surprises AI rivals with ultra-low cost Doubao modeltechnode.com

14. Mai 2024

GPT-4ononint.com
TechCrunch Minute: Everything you need to know about OpenAI’s new GPT-4o modeltechcrunch.com
GPT-4o delivers human-like AI interaction with text, audio, and vision integrationartificialintelligence-news.com
OpenAI’s ChatGPT announcement: Watch the GPT-4o reveal and demo heretechcrunch.com

13. Mai 2024

OpenAI debuts GPT-4o ‘omni’ model now powering ChatGPTtechcrunch.com

10. April 2024

OpenAI makes GPT-4 Turbo with Vision API generally availableartificialintelligence-news.com

07. März 2024

Why most AI benchmarks tell us so littletechcrunch.com

14. Februar 2024

What Is Generative AI?spectrum.ieee.org

06. Februar 2024

Need Help With Prompts? Ask me*community.openai.com

30. Dezember 2023

Need Help With Prompts? Ask me*community.openai.com

08. November 2023

Introducing Accurate Quantized Training (AQT) for accelerated ML training on TPU v5ecloud.google.com
Announcing Cloud TPU v5e GA for cost-efficient AI model training and inferencecloud.google.com

06. November 2023

OpenAI debuts GPT-4 Turbo and fine-tuning program for GPT-4techcrunch.com
As OpenAI’s multimodal API launches broadly, research shows it’s still flawedtechcrunch.com

19. September 2023

Need Help With Prompts? Ask me*community.openai.com

23. August 2023

OpenAI introduces fine-tuning for GPT-3.5 Turbo and GPT-4artificialintelligence-news.com

27. Juni 2023

Introducing AI Help (Beta): Your Companion for Web Developmentdeveloper.mozilla.org

23. Juni 2023

MosaicML’s latest models outperform GPT-3 with just 30B parametersartificialintelligence-news.com

13. Juni 2023

LLMs and Apple Vision: Exploring the Future of AI-Powered AR Experiencesallganize.ai

23. Mai 2023

Microsoft’s Azure AI Studio lets developers build their own AI ‘copilots’techcrunch.com

13. April 2023

Universal LLM Jailbreak: ChatGPT, GPT-4, BARD, BING, Anthropic, and Beyondadversa.ai

20. März 2023

AI Red Teaming LLM for Safe and Secure AI: GPT4 Jailbreak ZOOadversa.ai

30. November 2022

OpenAI upgrades GPT-3 with impressive new skillsartificialintelligence-news.com

27. Juli 2022

OpenAI’s GPT-3 is a convincing philosopherartificialintelligence-news.com

15. Dezember 2021

OpenAI now allows developers to customise GPT-3 modelsartificialintelligence-news.com

18. November 2021

OpenAI removes GPT-3 API waitlist and opens applications for all developersartificialintelligence-news.com

06. Juli 2021

Experts debate whether GitHub’s latest AI tool violates copyright lawartificialintelligence-news.com

06. Januar 2021

OpenAI’s latest neural network creates images from written descriptionsartificialintelligence-news.com

28. Oktober 2020

Medical chatbot using OpenAI’s GPT-3 told a fake patient to kill themselvesartificialintelligence-news.com

23. September 2020

Microsoft is granted exclusive rights to use OpenAI’s GPT-3artificialintelligence-news.com

10. September 2020

Expert calls out ‘misleading’ claim that OpenAI’s GPT-3 wrote a full articleartificialintelligence-news.com

01. Januar 2001

MLOpsairflow.apache.org
Modelldaten und Nachrichten aus BrunoSan AI-News, OpenRouter und Hugging Face. Benchmark-Werte stammen aus den jeweils verlinkten Quellen; Testbedingungen können abweichen. Modelle bleiben auch ohne Benchmark-Werte sichtbar.