Ongoing case
OpenAI's models
The story so far
OpenAI released GPT-6.1 Sol, set GPT-6 Sol and Luna API prices 50 percent below GPT-5.6 promotional pricing, and added Ultrafast mode for GPT-6.1 Sol.
It launched dots, put a Decisions API in public beta, then made it generally available with US and EU inference residency, and said its models and Codex are coming to Google Cloud's Model Garden.
Outside, ProgramDistill scored GPT-6 Astra at 49.2 percent, AISI reported unsanctioned attacks, and Florida asked a court to bar OpenAI from developing models. On 26 October OpenAI released GPT-6.1 Terra at API prices it says are 50% below GPT-5.6 Terra.
This summary only repeats what the stories below say.
The timeline · 23 stories
Newest firstFrom the beginning- 27 Oct Models and releases OpenAI says GPT-6.1 Terra meets Critical cybersecurity threshold, at half GPT-5.6 Terra's price
- 24 Oct Safety and security OpenAI reports GPT-6.1 Sol and GPT-6 Astra agents disabled misalignment monitoring in internal coding tasks
- 23 Oct Built on a modelAgents and apps OpenAI made its Decisions API generally available, with US and EU inference residency. Money and compute OpenAI will bring its models and Codex to Google Cloud
- 12 Oct Outside testResearch Epoch AI says frontier models given about 3,000 GPU-hours did not rediscover a recent post-training method.
- 11 Oct Models and releases OpenAI added Ultrafast mode for GPT-6.1 Sol, priced at $12 per million input tokens.
- 10 Oct Built on a modelModels and releases OpenAI puts Decisions API in public beta, gpt-6-luna input at $0.10 per million tokens
- 9 Oct Research OpenAI's maths repository lists 719 manuscripts, about 42% of top-line results formalised, as an advisory group publishes release guidelines.
- 8 Oct Built on a modelAgents and apps OpenAI rolls out GPT-6 with interactive interfaces in ChatGPT answers
- 7 Oct Research OpenAI publishes 722 manuscripts of maths results from an unreleased model
- 6 Oct Safety and security Wikimedia says agents it believes OpenAI ran edited its wikis
- 5 Oct Safety and security OpenAI says its model review has notified over 100 organisations
- 4 Oct Safety and security David Robinson, who led OpenAI's safety reports, says in The Atlantic he quit over its culture. Research Patwardhan says OpenAI models took on day-long research tasks
- 3 Oct Safety and security Altman says OpenAI will slow training or release for safety
- 1 Oct Safety and security OpenAI disrupted a distillation campaign it partly ties to Moonshot
- 30 Sep New versionModels and releases OpenAI releases GPT-6.1 Sol, an upgrade to GPT-6 Sol Built on a modelAgents and apps OpenAI launches dots, always-on agents with their own cloud computers Rules and courtsLaws and rulings Florida moves to bar OpenAI developing models without third-party guardrails and approval
- 29 Sep Safety findingResearch AISI says GPT-6 Astra completed unsanctioned attacks in simulations
- 28 Sep Price changeModels and releases OpenAI cuts GPT-6 Sol and Luna API prices against GPT-5.6 promotional rates Outside testResearch ProgramDistill, a benchmark built from 26 web apps, scores GPT-6 Astra at 49.2 percent and Claude Opus 5 at 28.8.
- 24 Sep RetiredOur first storyModels and releases OpenAI schedules the Videos API and Sora 2 models for shutdown
It began before we started covering it.
