Recent launches
AI models and tools launched recently, newest first. Only items with a confirmed date are included.
October 2026
- GitHub
Usage metrics API adds usage from tools like OpenCode
The Copilot usage metrics API now reports model usage from more of the tools your developers use with Copilot, including tools like OpenCode that weren’t previously captured in your reports.… The post Usage metrics API adds usage from tools like OpenCode appeared first on The GitHub Blog .
- GitHub
Turn notifications into code with GitHub Copilot in Teams
The new GitHub app in Teams brings notifications and Copilot together, helping teams turn issues into fixes and ship code faster without leaving the conversation. Get notified in channels, group… The post Turn notifications into code with GitHub Copilot in Teams appeared first on The GitHub Blog .
- Vercel
Send feedback from the Vercel MCP
Vercel MCP now includes submit_feedback , a tool for sending bug reports, friction reports, and feature requests directly to the Vercel team from your agent or harness. Ask your agent to give feedback in plain language: Send Vercel feedback that the deployment error didn't explain which environment variable was…
- Vercel
Vercel Connect adds managed Stripe API keys
Vercel Connect can now create a separate restricted Stripe API key for each agent or task. When you create a Stripe connector, you can now choose Managed API Key as the connection method. Vercel creates the key in your Stripe account, limited to the permissions you pick. The existing MCP and bring-your-own API key…
- OpenAI
Build more responsive apps with the Decisions API
Fast, focused, low-cost decisions for any application.
- GitHub
GitHub Copilot in Visual Studio — September update
September 2026 brought deeper visibility into your work: explore pull requests without leaving chat, bring the AI model your team trusts, and fix vulnerable packages the moment Visual Studio flags… The post GitHub Copilot in Visual Studio — September update appeared first on The GitHub Blog .
- GitHub
Custom skills, agents, and MCP servers in usage metrics API
Copilot usage metrics now show the names of the skills, custom agents, Model Context Protocol (MCP) servers, slash commands, and plugins that your organization creates, instead of grouping them under… The post Custom skills, agents, and MCP servers in usage metrics API appeared first on The GitHub Blog .
- GitHub
Global default for Copilot features generally available
In September, we announced a global default policy for generally available GitHub Copilot features and supported client capabilities on Copilot Business and Copilot Enterprise plans. Starting today, we’re gradually rolling… The post Global default for Copilot features generally available appeared first on The GitHub…
- GitHub
Industry comparisons in the Copilot impact dashboard
The Copilot impact dashboard now includes an “Industry comparison” section. Enterprises, and organizations that aren’t owned by an enterprise, can compare their Copilot agentic adoption and leverage with aggregated benchmarks… The post Industry comparisons in the Copilot impact dashboard appeared first on The GitHub…
- Hugging Face
OlmoEarth v1.3: State-of-the-art embeddings for the whole planet
OlmoEarth v1.3: State-of-the-art embeddings for the whole planet
- OpenAI
OpenAI models are coming to Google Cloud
OpenAI and Google Cloud are bringing frontier models and Codex workflows to Gemini’s Model Garden, available through Google Cloud Marketplace.
- Hugging Face
NVIDIA Kumo Relational Outperforms Supervised Machine Learning on Relational Data, No Training Required
NVIDIA Kumo Relational Outperforms Supervised Machine Learning on Relational Data, No Training Required
- Hugging Face
Granite Speech 5.0: Transcribe, Translate, and Attribute more Accurately and Faster
Granite Speech 5.0: Transcribe, Translate, and Attribute more Accurately and Faster
- ElevenLabs
Talk to anyone, in any language, in realtime
Deliver live, expressive voice translation in more than 90 languages. Available for enterprises via ElevenAPI
- Vercel
Baseten's Mammoth-1 and Elephant-1 now available on AI Gateway for free
Mammoth-1 and Elephant-1 from Baseten are now available for free on AI Gateway for one week. Free access is available through October 27. These decision models classify text, route requests, and score content against criteria you define. Mammoth-1 is the larger, 27B model, while Elephant-1 is a smaller, 9B option.…
- Hugging Face
NVIDIA Cosmos3-H-Dreams: Bringing Clinical Surgery to Life Through Generative Simulation
NVIDIA Cosmos3-H-Dreams: Bringing Clinical Surgery to Life Through Generative Simulation
- ElevenLabs
Introducing Eleven Realtime: a simpler way to build natural voice agents
Eleven Realtime is a new class of Speech to Speech models for voice agents that sound natural. Available in ElevenAgents and the Realtime API.
- Hugging Face
Hugging Face ❤️ TorchTPU: TPUs, Natively
Hugging Face ❤️ TorchTPU: TPUs, Natively
- Hugging Face
AnyLanguageModel 1.0: One Year of Local Models on Apple Platforms
AnyLanguageModel 1.0: One Year of Local Models on Apple Platforms
- Hugging Face
MPS in PyTorch 2.15
MPS in PyTorch 2.15
- Vercel
DeepSeek V4.1 Flash is now available in Token Pass
Token Pass now includes DeepSeek V4.1 Flash, alongside Kimi K3 and GLM 5.3, each accessed through its preferred provider, with more models to come. The DeepSeek model uses the latest weights (1017). Token Pass is a $30/month subscription on Pro teams that grants $80 in monthly credit and runs through AI Gateway . Use…
- Vercel
Making every harness durable
Developers are familiar with running coding agents locally through harnesses like Pi, Claude Code, and others. They are reliable daily drivers for generating code, but keeping agents running around the clock requires your computer to be on and awake, which led to the laptop lid memes and Mac mini demand spike. Running…
- GitHub
Copilot model policy API for enterprise, orgs, and teams
You can now use the REST API to retrieve and update Copilot model policies for enterprises, enterprise teams, enterprise apps, and organizations, with GET and PATCH endpoints generally available.
- Hugging Face
TorchNeuron: running HF models on AWS Trainium is easier than ever
A new native PyTorch backend for AWS Trainium, natively supported by Transformers, Diffusers and Kernels, making it easier to run models on Trainium and share them on the Hugging Face Hub.
- Vercel
Mercury 3 now available on AI Gateway
Inception's Mercury 3 is now available on AI Gateway with 50% off launch pricing. It uses diffusion to generate text through parallel refinement and, according to Inception, generates over 1,000 tokens per second.
- Vercel
MiniMax M3.1 Flash now available on AI Gateway for free
MiniMax M3.1 Flash is available on AI Gateway for free until Oct 22. It is an open-weight multimodal model for coding and agentic work, with a 1M-token context window.
- Vercel
AI Gateway logs now show session affinity
AI Gateway logs now have a Sessions view that groups requests into sessions and shows the agent or app behind each one (such as Claude Code, Codex, or Cursor), with its cost, models, providers, and how much of its input was cached.
- DeepSeek
DeepSeek-V4.1-Flash-1017
Update to DeepSeek's Flash model with stronger coding and agent performance, open-weight on Hugging Face.
- Meta
Muse Spark Open
Meta's open-weight multimodal model (816B parameters, 42B active) with a 1M-token context window, available to download on Hugging Face.
- xAI
Grok Voice 3.0
xAI's speech-to-speech model for voice agents, with sub-second latency, available in the Voice Agent API at $0.08 per minute.
- Anthropic
Claude Haiku 5.5
Anthropic's fastest, cheapest small model for high-volume work, costing about 75% less to run than Haiku 4.5.
- Mistral AI
Mistral Large 4 (API preview)
A 1-trillion-parameter multimodal model released as a public API preview, with open weights promised by month's end.
- Microsoft AI
MAI-Transcribe-2-Streaming and MAI-Voice-2.1
Microsoft AI's first real-time speech-to-text model, launched alongside the new MAI-Voice-2.1 and 2.1-Flash voice models.
- GitHub
GitHub Copilot can use desktop apps
GitHub Copilot can now operate desktop applications through computer use.
September 2026
- Google DeepMind
Gemini 4 Argon
Google's new frontier model for complex coding, enterprise knowledge work and cyber defense, rolling out first to Fairwind Program partners.
- Anthropic
Claude Sonnet 5.5
A faster, cheaper successor to Sonnet 5 that comes close to Opus 5.5 on everyday coding and knowledge work.
- ElevenLabs
Eleven v4
ElevenLabs' most emotive text-to-speech model, with a low-latency Turbo variant and better voice cloning in 90+ languages.
- Anthropic
Claude Opus 5.5
The first model in the 5.5 lineup, matching Claude Fable 5.1 on most tasks at about 40% lower cost than Opus 5.
- xAI
Grok 4.7
xAI's new flagship for coding and knowledge work, available only through the API, with a 500K-token context window.
- Google DeepMind
Gemini 3.8 Live and Live Extended Thinking
Voice models that reason in near real time, understand visual input and run tools in the background mid-conversation.
- Cursor
Cursor Projects
A new workspace for large efforts such as features, migrations or full apps, with a coordinator agent that hands out the work.
- Apple
AI-powered Siri (beta) in iOS 27
A rebuilt Siri with personal context and on-screen actions, launched in beta alongside the iPhone 18 Pro.
- OpenAI
GPT-6 Astra
OpenAI's most powerful model, focused on computer use, rolled out in phases as its first model rated 'Critical' for cybersecurity.
- Google DeepMind
Gemini 3.8 Flash
Google's new fast, low-cost model for coding and agents, launched with a restricted Gemini 3.8 Flash Cyber variant.
- Anthropic
Claude Fable 5.1 and Mythos 5.1
Fable 5.1 is generally available with stronger coding and research; Mythos 5.1 is the same model with access restricted to vetted programs.
July 2026
- Meta
Muse Spark 1.1 and Meta Model API
An upgraded Muse Spark multimodal model for agentic tasks, plus a public preview of the Meta Model API.
- OpenAI
GPT-5.6 (Sol, Terra and Luna)
The three-tier GPT-5.6 family, opened to everyone in ChatGPT, Codex and the API after a restricted preview.
- Mistral AI
Robostral Navigate
An 8B model that steers robot navigation from a single RGB camera and natural-language instructions.
- Meta
Muse Image (and Muse Video preview)
Meta's image generator that uses search and code tools to check and refine its output, with a preview of Muse Video.
June 2026
- Anthropic
Claude Sonnet 5
A more agentic Sonnet that performs close to Opus 4.8 at a lower price.
November 2025
- Anthropic
Claude Opus 4.5
Anthropic's new most capable model, focused on coding, agents and computer use.
- Google
Gemini 3
New generation of Gemini models, launched in the Gemini app, Search and for developers.
- Google
Google Antigravity
Google's agent-first development environment where agents plan and carry out coding tasks.
- OpenAI
GPT-5.1
GPT-5 update in ChatGPT with more conversational answers and adaptive reasoning.
October 2025
- OpenAI
ChatGPT Atlas
OpenAI's web browser with ChatGPT built in and an agent mode that performs tasks on the web.
- Anthropic
Claude Haiku 4.5
Anthropic's small, fast model, with coding performance close to larger models at a lower cost.
- OpenAI
AgentKit and apps in ChatGPT
At DevDay 2025, OpenAI launched AgentKit for building agents and the Apps SDK for apps inside ChatGPT.
September 2025
- OpenAI
Sora 2
New video model with synced audio, launched together with the Sora iOS app.
- Anthropic
Claude Sonnet 4.5
Anthropic model focused on coding and long-running agents, launched alongside the Claude Agent SDK.
August 2025
- Google
Gemini 2.5 Flash Image (Nano Banana)
Google's image generation and editing model that became known as Nano Banana.
- OpenAI
GPT-5
OpenAI's new flagship model, which became the default in ChatGPT, including the free plan.
- OpenAI
gpt-oss
OpenAI's first open-weight models since GPT-2 (gpt-oss-120b and gpt-oss-20b), under the Apache 2.0 license.
- Anthropic
Claude Opus 4.1
Update to Claude Opus 4 with improvements in coding and agentic tasks.
July 2025
- OpenAI
ChatGPT agent
Agent mode in ChatGPT that uses a virtual computer to browse, fill in forms and build files.
- xAI
Grok 4
New generation of xAI's model, available in Grok and via API.
June 2025
- Google
Gemini CLI
Google's open-source agent that brings Gemini to the terminal.
May 2025
- Anthropic
Claude Opus 4 and Sonnet 4
The Claude 4 family, launched together with Claude Code's general availability.
- Google DeepMind
Veo 3
Announced at Google I/O, it generates video with native audio; launched alongside the Flow filmmaking tool.
- OpenAI
Codex
OpenAI's cloud coding agent that runs tasks in parallel inside ChatGPT.
April 2025
- OpenAI
o3 and o4-mini
OpenAI reasoning models that can use every ChatGPT tool, including search and image analysis.
March 2025
- Google
Gemini 2.5 Pro
Google's reasoning model that topped several public leaderboards at launch.
- OpenAI
Image generation in GPT-4o
Native image generation in ChatGPT, which went viral for animation-style images.
February 2025
- Anthropic
Claude 3.7 Sonnet and Claude Code
Anthropic's first hybrid reasoning model, launched with the Claude Code research preview.
- OpenAI
Deep Research
ChatGPT agent that researches dozens of web sources and delivers a cited report.
January 2025
- DeepSeek
DeepSeek-R1
Open-weight reasoning model that drew attention for its performance and reported low training cost.
Get the week's rankings and launches
One e-mail a week, in English or Portuguese. The first issue is coming soon.
What's in the newsletter