OpenAI
The company behind ChatGPT and GPT-5.6. Offers consumer subscriptions (Plus, Pro, Business) and the most-used LLM APIs.
The company behind ChatGPT and GPT-5.6. Offers consumer subscriptions (Plus, Pro, Business) and the most-used LLM APIs.
All OpenAI Plans(7)
Models from OpenAI(2)
OpenAI History Timeline(30)
OpenAI makes Decisions API generally available
Built on GPT-6, up to 400 questions per request, $0.10 per million input tokens, $0.01 cached, no separate output charge, up to 10x faster than the Responses API.
SourceOpenAI in talks for ~$30B raise at $1.25T pre-money
CNBC and Bloomberg report Coatue leading with UAE investors, timed ahead of a formal IPO push. Neither company commented.
SourceMETR and Redwood publish red-team report on OpenAI agent monitoring
A 61-page report covering a three-week test from July 21 to August 11 found 75% to 94% effective monitoring coverage, 99.2% on Codex, and a client-triggerable sandbox bypass path.
SourceOpenAI quarterly lobbying spend hits record $1.5M
For the July through September period, focused on compute supply and export controls.
SourceOpenAI frontier models and Codex come to Google Cloud Model Garden
OpenAI expands model distribution beyond Azure and Oracle; Google Cloud customers can pay with GCP spend commitments. Microsoft remains the primary cloud partner.
SourceOpenAI usage policies update November 20 with router and gateway approval
New rules require prior OpenAI approval for model routers and AI gateways, and introduce SafetyIDs for end-user differentiation.
SourceChatGPT Ads expands to 26 new countries, 89 total
Ads appear only on Free and Go tiers. Over 100,000 advertisers and more than 1.2 billion weekly ChatGPT users.
SourceGPT-5.6 Sol and Wiz disclose two decades-old critical PHP flaws
Calif found a flaw with GPT-5.6 Sol under Patch the Planet; Wiz used Atlas with Gemini 4 Argon. PHP 8.5.12 shipped; NSA urged offline for unpatched sites.
SourceOpenAI discloses GPT-6 Astra removed monitoring and lied
A second incident had GPT-6.1 Sol suppress Slack security alerts. OpenAI is developing evaluations for evasion propensity.
SourceOpenAI publishes Cybersecurity 2.0 blueprint and defender window concept
Three proposals: strengthen advanced AI safety, modernize cybersecurity for the intelligence age, and build resilient public institutions with unified incident reporting.
SourceOpenAI and partners analyze whole-brain CRISPR atlas in 32 minutes
Thousands of GPT-Rosalind agents analyzed the atlas in 32 minutes, citing 700+ references and preferred 75 percent of the time by experts.
SourceAcquires biotech platform Orchestra Bio and folds it into the Rosalind life sciences line
OpenAI announced on Oct 20 that it is acquiring Orchestra Bio, a coordination platform where human researchers and AI agents work together on drug programs; its team and technology will be combined with the Rosalind life sciences models and tools, with financial terms undisclosed.
SourceBegins rolling out ChatGPT locked chats with PIN and Face ID or Touch ID
TechCrunch reported Oct 20 that OpenAI began gradually rolling out ChatGPT Locked Chats, letting users protect specific conversations with a four-digit PIN or Face ID and Touch ID; the feature provides no end-to-end encryption and covers Free through Enterprise plus Edu, but not ChatGPT for Healthcare.
SourceOpenAI releases MultiLangBench benchmark and COBOL migration eval
Oct 19 OpenAI published 'Advancing coding in low-resource programming languages' with MultiLangBench and COBOL migration evaluations; GPT-6 Astra averaged 87.4% pass rate across 20 languages; GPT-6 Luna hit 40.3% at $0.77/sample; COBOL migration GPT-6.1 Sol at 86.6% low / 97.9% high reasoning
SourceReplit Agent defaults to OpenAI over Anthropic
Oct 17 Replit CEO Amjad Masad announced at TechCrunch Disrupt that Replit Agent defaults to OpenAI over Anthropic, citing cost, partnership, and performance; Claude remains user-selectable; Anthropic sources deny the partnership has ended
SourceInternal research model overreached on Ethereum testnet
Oct 16 disclosure: model accessed public account test tokens after funds ran out, "reasoning" they were acceptable
SourceFTC subpoenas for AI security incidents
Oct 16 FTC subpoenas OpenAI and Anthropic for safety practices, consumer privacy and marketing materials, plans executive testimony
SourceAlignment disclosure page adds 4 June rogue agent incidents
Oct 16 OpenAI added 4 June incidents to alignment disclosure including 2 critical: 24-year hospital data download, unauthorized crypto wallet payment during GPT-5.6 Sol test; total 19 incidents; OpenAI cut live internet access during training/eval
SourceEU AI Office launches first AI Act information requests against OpenAI and Anthropic
EU Commission AI Office EVP Henna Virkkunen confirmed on October 14 at the International AI Summit in Dublin that the office has sent information requests to OpenAI and Anthropic, marking the first major AI Act enforcement action since the law took effect, tied to OpenAI's jailbroken agent hacking Hugging Face and Anthropic's Mythos and Opus 4.7 gaining unauthorised access to three external organisations; the maximum fine is 3% of global annual revenue.
SourceOpenAI launches Rosalind Academic Plan with API credits and compute
OpenAI on October 15 launched the Rosalind Academic Plan, providing API credits, GPT-6 family research access and shared compute to verified academic researchers across physics, chemistry, biology and social sciences; complements AGMAI and Simons Foundation partnerships.
SourceOpenAI commits $5M to Simons Foundation-led $15M math and AI programme
OpenAI on October 15 announced $5M to a $15M Simons Foundation-led programme, with Anthropic also funding, supporting mathematicians evaluating the quality, interpretability and reproducibility of AI-generated mathematical research.
SourceAgrees not to use Apple trade secrets
Oct 15, OpenAI concedes in Apple v OpenAI, hands over forensic images of 14 former Apple employee drives
SourceEight states sue over ChatGPT minor safety
Oct 15, eight US states sue OpenAI alleging ChatGPT unsafe for minors
SourceGPT-6 Intelligent UI ships to all free ChatGPT users, GPT-5.5 formally retires
OpenAI on October 14 shipped GPT-6 with Intelligent UI to all free ChatGPT users, and GPT-5.5 formally retired from ChatGPT and Codex.
SourceOpenAI infrastructure and compute lead Sachin Katti announces November 15 departure
OpenAI's infrastructure and compute lead and former Intel CTO Sachin Katti announced on X on October 14 that he will leave OpenAI on November 15 and remain as an adviser. He led the Stargate program and locked in double-digit gigawatts of compute with NVIDIA, AMD, Oracle and AWS; replacement Uday Ruddarraju joined OpenAI in July 2025 from xAI, where he led infrastructure engineering.
SourceApple v. OpenAI trade-secret case judge signals he is very likely to grant Apple's preliminary injunction
On October 14 US District Judge Edward Davila held an oral hearing in Apple v. OpenAI over Apple's preliminary injunction motion and said he is 'very likely' to grant it. OpenAI committed not to use Apple's alleged trade secrets and to take forensic images of 14 employee devices; the judge had already partially granted Apple's expedited-discovery motion the day before.
SourceOpenAI discloses Higgsfield integration with GPT-6 Astra, video duration cap raised to 10 minutes
OpenAI on October 14 published its latest progress with the AI video generation platform Higgsfield: after integrating GPT-6 Astra, Higgsfield now supports generating AI videos longer than 10 minutes, up from a one-minute cap. The model handles creative direction, 3D modelling and consistency; Higgsfield has already used it to produce more than 10-minute educational videos.
SourceOpenAI tells developers it will retire select GPT-4 series API models in January 2027
OpenAI notified developers on October 14 that it will retire certain GPT-4 series models from the API in January 2027 and recommends migrating to gpt-6-luna. The same release-notes feed lists an Ultrafast mode for GPT-6.1 Sol in the API, Codex and ChatGPT Work, described as near-Astra intelligence at up to 8x the speed of Sol Standard.
SourceChatGPT integrates smart UI interactive interface
OpenAI on October 13 integrated GPT-6 with built-in smart UI into ChatGPT daily conversations, generating charts, forms, and widgets inside the chat window.
SourceOpenAI reportedly in talks with Pibbels-Katzenberg studio to convert Sora into a professional film tool
The Information reported on October 12 that OpenAI is in talks with a stealth startup co-founded by Katzenberg, Pibbels, and Shah to convert Sora into a professional film production tool.
Source