New AI models and releases. Newest first, every story linked to its sources.
Meta introduced Muse Wonder and Muse Wonder Full for agentic work, coding, and expert domains. Wonder is available via the Meta Model API and Muse Code; Wonder Full is limited to Muse Code subscribers.
Microsoft introduced Maia 300, the third generation of its in-house AI accelerator, and said it will scale the Azure Maia family to multi-gigawatt deployments starting in 2027 as its primary inference system.
OpenAI has begun a wide commercial rollout of GPT-5.6 in Luna, Terra and Sol tiers, paired with ChatGPT Work, a mode for long-running background tasks, after a period of staggered access and Commerce Department review.
Anthropic's new dynamic workflow feature lets a main Claude agent write a program that splits work across as many as 1,000 sub-agents. In an Anthropic test, it found 66 of 70 planted bugs in a 116,000-line codebase.
Odyssey made its Odyssey-3 generative world model free to try as a public research preview. The company says it builds interactive environments from text in real time and can control robots, drones and a running copy of GTA V.
Microsoft released Microsoft-Decision-1, a model that scores options for decisions such as routing, classification and verification rather than writing text. It is available in Microsoft Foundry and on OpenRouter.
OpenAI began rolling out a GPT-6 model in ChatGPT that can answer with interactive interfaces. Paid tiers got it first, with free users following the next day.
Mistral opened a public preview of Mistral Large 4, a natively multimodal model with about 1 trillion total and 52 billion active parameters. The French lab says the weights will be released by the end of October.
Google DeepMind released EmbeddingGemma 2, a 740-million-parameter open model that maps text, code, images, audio and video into one shared embedding space. It is released under the Apache 2.0 license.