AI model releases
New AI models, tests, and leaks, newest first. Each line links to the original source. Rumors and leaks are in the list, and their headlines say so.
Latest by lab
- Microsoft Microsoft releases Aion Instruct, a 2.3B open-weights model that replaces Phi-4-mini as the built-in AI in Edge Nov 20
- Google Google Genie 4 is now available to developers in the Gemini API and AI Studio, generating 720p worlds at 24fps Nov 20
- OpenAI OpenAI releases GPT-Live-1.1 in the API with more natural accents, a standby option, and more reliable task completion Nov 19
- Anthropic Microsoft releases MAI Code 2 and MAI Code 2 Flash coding models, matching 96 percent of Claude Opus 5.5 on its everyday coding average at 5 percent of the cost Nov 19
- Alibaba Alibaba releases Qwen4-Max, a 2.4 trillion parameter model with a 1M token context window and API prices 62 percent cheaper per task than Qwen3.8-Max Nov 15
All model news
November 2026
-
Thinking Machines releases Inkling 2 Mini, a 26B open-weights AI model that runs on a laptop and loops its middle layers to reason
Inkling 2 Mini is an open-weights model small enough to run on your laptop. It loops its middle layers to reason more per token and reach answers in fewer tokens.
-
Microsoft releases Aion Instruct, a 2.3B open-weights model that replaces Phi-4-mini as the built-in AI in Edge
Rapidly evolving AI capabilities are redefining what sites can do for users on the web. With on-device AI in Microsoft
-
Microsoft launches MAI Realtime 1, its first speech-to-speech model, at $0.01 per audio minute
We make responsible AI to empower people's lives.
-
Google Genie 4 is now available to developers in the Gemini API and AI Studio, generating 720p worlds at 24fps
Build interactive 720p worlds with Genie 4 on the Google Gemini API. Generate coherent environments with native audio and real-time controls. Get started.
-
OpenAI releases GPT-Live-1.1 in the API with more natural accents, a standby option, and more reliable task completion
-
Introducing Tool Retrieval in the Gemini API
Give your agents access to hundreds of tools while using fewer tokens with Tool Retrieval, now in preview in the Gemini API.
-
Microsoft releases MAI Code 2 and MAI Code 2 Flash coding models, matching 96 percent of Claude Opus 5.5 on its everyday coding average at 5 percent of the cost
-
A leak says Google is preparing to release the Gemma 5 AI model, with no official announcement
-
Google releases Gemini 4 Neon, sitting between Carbon and Hydrogen, in the Gemini API, AI Studio and the Gemini app
Bringing frontier-level intelligence with fast end-to-end task completion across complex software engineering, knowledge work, and science workflows.
-
Microsoft launches MAI Cyber 2 Flash security model
MAI Cyber 2 Flash scores 92.3% on CyberGym-E2E with a one-million-token context and arrives exclusively in Microsoft security products like Codename MDASH.
-
OpenAI releases GPT-6.2 Astra and Sol
-
’Sleeper Agent’ Tactic Hides Payloads in Open-Weight AI Models
-
Liquid AI releases d1.2 and d1.2-ultra, omnimodal decision models that process text, images, and audio
-
Alibaba releases Qwen4-Max, a 2.4 trillion parameter model with a 1M token context window and API prices 62 percent cheaper per task than Qwen3.8-Max
Qwen offers comprehensive functionality spanning chatbot, image and video understanding, image generation, document processing, web search integration, tool utilization, and artifacts.
-
Google launches Gemini 4 Hydrogen, a fast coding and agent model that edges out Gemini 4 Argon, rolling out across Gemini tiers at 1.50 dollars per million input tokens
Today, we’re introducing Gemini 4 Hydrogen, our daily driver for agentic workflows
-
Google pushes privacy angle as Gemini-powered smartglasses debut starting at $395
The first Gemini smartglasses are a collab between Google, Samsung, Gentle Monster, and Warby Parker.