Remonter à la source.
Comprendre ce qui change. Choisir ce qui mérite un test.
Notre méthode
Chaque fiche distingue les faits annoncés, la disponibilité réelle, les limites et une recommandation éditoriale. Les démonstrations officielles illustrent une annonce ; elles ne remplacent pas un test indépendant.
Le journal suit l’actualité de l’IA semaine après semaine. Chaque article renvoie à ses sources et distingue les annonces, la disponibilité réelle et les points encore incertains.
Une sélection éditoriale, pas un inventaire exhaustif de toutes les annonces IA. Les références ci-dessous accompagnent les articles des archives.
Découverte
Les médias spécialisés et newsletters complètent les annonces officielles, les documentations techniques et les publications de recherche.
OHMO ↗, Evolving AI Insights, Melvynx / Le Signal IA, newsletters fournisseurs et développeurs.
Sources primaires
- Release notes | Gemini API | Google AI for Developers ↗
- Gemini 4 Argon is now available in the Gemini app ↗
- Gemini 4 Argon: our next era of frontier intelligence ↗
- Gemini API : régions disponibles ↗
- Developing Gemini 4 Argon responsibly and safely ↗
- Introducing Claude Haiku 5.5 \ Anthropic ↗
- Claude Platform release notes - Claude Platform Docs ↗
- Muse Wonder: A Frontier Model For Personal & Professional AI Agents ↗
- Meta Newsroom : Introducing Muse Wonder models ↗
- ChatGPT : notes de version ↗
- GPT-6 Sol and GPT-6 Luna: October 2026 update - OpenAI Deployment Safety Hub ↗
- OpenAI DevDay : récap officiel ↗
- Open From Data Center to Desktop: NVIDIA Expands Nemotron Family With Highest Efficiency Super VL and Bolt Models | NVIDIA Blog ↗
- nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-NVFP4 · Hugging Face ↗
- nvidia/NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB · Hugging Face ↗
- Build live dashboards and animate explainers with Claude | Claude by Anthropic ↗
- GPT-6.1 Sol : documentation API ↗
- OpenAI API : changelog ↗
- Gemini 3.8 : génération vocale ↗
- Introducing Muse Spark Open, The Most Capable US Open‑Weight Model to Date | Meta AI Research ↗
- meta-models/Muse-Spark-Open-816B-A42B · Hugging Face ↗
- Introducing Grok Voice 3.0 ↗
- Claude Sonnet 5.5 : annonce officielle ↗
- Claude Opus 5.5 : annonce officielle ↗
- Gemini Live avec Live Avatar ↗
- AI Scan for pull requests generally available - GitHub Changelog ↗
- Agentic autofix for code scanning alerts general availability - GitHub Changelog ↗
- Introducing Mistral Large 4 | Mistral ↗
- Customize Claude Code with mods ↗
- anthropics/claude-code v2.1.287 ↗
- Claude Resources ↗
- Claude et la découverte du système ART ↗
- Trace agents with OpenTelemetry on the Claude Platform ↗
- ChatGPT & Codex changelog | ChatGPT Learn ↗
- Introducing Clef: our open-source decision models, and new RL fine-tuning platform | Cloudflare Blog ↗
- AI models may be capable of self-replicating without access to a datacenter \ Anthropic ↗
- Claude now works with Google Docs, Sheets, and Slides | Claude by Anthropic ↗
- Our first streaming transcription model debuts at no. 1 on Artificial Analysis | Microsoft AI ↗
- Introducing Beam: Reflection’s 501B open-weight model : Reflection ↗
- Muse Gadgets: Open source hardware for your Muse ↗
- Meta wants your next gadget to be Muse-infused | TechCrunch ↗
- Our approach to EU text provenance rules | OpenAI ↗
- Monthly API credits for Max and Team plans | Claude Help Center ↗
- Change Log | DeepSeek API Docs ↗
- Gemini at Work 2026: Introducing Gemini agent | Google Cloud Blog ↗
- Introducing Hark Pro | Hark ↗
- Anthropic update on independent evaluators \ Anthropic ↗
- Designing more efficient quantum circuits with Claude \ Anthropic ↗
- Changelog Oct 16, 2026 - Neon ↗
- Character.ai Initiates Voluntary Chapter 11 Proceeding to Facilitate Value-Maximizing Sale Process ↗
- Code Review is now generally available | Claude by Anthropic ↗
- Search & Fetch | DeepSeek API Docs ↗
- Frontier Pre-training at Microsoft AI | Microsoft AI ↗
- Meet Apple HomeView, a new way to enjoy entertainment and manage the home - Apple ↗
- Introducing The X Subscription | SpaceXSI ↗
- Cloud Agent Canvases and Environment Improvements · Cursor ↗
- Lovable changelog - Lovable Documentation ↗
- ColGrep-Agent: Cutting Frontier Coding Agents’ Tokens by a Third ↗
- Claude reveals ECDSA curve origins \ Anthropic ↗
- Bringing GPT-Rosalind to academic research labs | OpenAI ↗
- How employment is changing at Anthropic \ Anthropic ↗
- Introducing DigitalOcean Agent Data Services: The Data Foundation for Agents | DigitalOcean ↗
- DigitalOcean Inference Router Enhancements: Performance Tuned Routing, No Rules to Write | DigitalOcean ↗
- Introducing DigitalOcean Signals and Simulations: see how your agents and models really behave in production, then test the fix before it ships | DigitalOcean ↗
- The AI expectations gap \ Anthropic ↗
- CERN and Mistral join forces to accelerate AI for science – Home | CERN ↗
- API Changelog & Updates | Runway Dev ↗
- Investing in the public benefit \ Anthropic ↗
- Google’s new Universal Cart launches today in the U.S. ↗
- GitHub Actions workflow lockfiles in public preview - GitHub Changelog ↗
- Open source license compliance generally available - GitHub Changelog ↗
- AI credits - Mobbin Docs ↗
- Upcoming deprecation of selected GitHub Copilot models in mid-November - GitHub Changelog ↗
- Multi-repository support and other improvements for GitHub in Slack and Teams - GitHub Changelog ↗
- Higgsfield takes AI video from 1 minute to over 10 | OpenAI ↗
- Tokenmaxxing BERT: 50x faster tokenization ↗
- Halloween Marketing Ideas: Spooky AI Videos | Artlist — Artlist Blog ↗
- Artlist Blog ↗
- anthropics/claude-code v2.1.299 ↗
- anthropics/claude-code v2.1.300 ↗
- Preparing for the 2026 US midterms \ Anthropic ↗
- HydraFusion updates and new Max option - GitHub Changelog ↗
- The EU’s next chapter in AI is agentic | OpenAI ↗
- Origin App Marketplace · Cursor ↗
- All Systems Go: Global PC App Ecosystem Ready for NVIDIA RTX Spark | NVIDIA Blog ↗
- Strengthening our defenses against distillation attacks \ Anthropic ↗
- Special meeting of the AI Scientific Panel | Shaping Europe’s digital future ↗
- Investigating unintended model actions in our evaluations and internal use \ Anthropic ↗
- Changes to Gemini model access and limits - Gemini Apps Help ↗
- Google Downgrades Every Free Gemini User ↗
- anthropics/claude-code v2.1.295 ↗
- Introducing the Anthropic Cyber Mission \ Anthropic ↗
- An opt-in vulnerability-finding service for open-source software \ Anthropic ↗
- openai/codex rust-v0.162.0 ↗
- openai/codex rust-v0.161.0 ↗
- openai/codex rust-v0.162.1 ↗
- ChatGPT Enterprise and Edu release notes | OpenAI Help Center ↗
- Disrupting AI-enabled “false front” operations | OpenAI ↗
- Discover local models in GitHub Copilot CLI - GitHub Changelog ↗
- Local sandboxing for GitHub Copilot now generally available - GitHub Changelog ↗
- NVIDIA, Microsoft Kick Off a New Beginning for Windows PCs with RTX Spark and AI Agents | NVIDIA Blog ↗
- Helping teens learn, plan, and shape the future of AI | OpenAI ↗
- Expanding the Cyber Verification Program \ Anthropic ↗
- Remote control for local agents · Cursor ↗
- Bring multimodal semantic search to the edge with EmbeddingGemma 2 - Google Developers Blog ↗
- Rate limits | OpenAI API ↗
- Decisions | OpenAI API ↗
- Sharing AI progress in mathematics | OpenAI ↗
- Building advertising for the way people use AI | OpenAI ↗
- NVIDIA DGX Spark 64GB Gives Developers More Ways to Build and Scale Local AI | NVIDIA Blog ↗
- GitHub Copilot can now interact with desktop apps with computer use - GitHub Changelog ↗
- Dynamic workflows in Copilot CLI and the Copilot app - GitHub Changelog ↗
- openai/codex rust-v0.160.0 ↗
- GLM-5.3 : évaluation Anthropic ↗
- GLM-5.3 : évaluation NIST / CAISI ↗
- Ollama : modèles de décision ↗
- Project Suncatcher : tests et défis ↗
- Cursor Projects : annonce officielle ↗
- Meta Connect : récap officiel ↗
- Claude Code Projects : documentation ↗
- Singapore office and government MOU | Anthropic \ Anthropic ↗
- Building on our commitment to American scientific discovery \ Anthropic ↗
- 2026 Usage Policy update \ Anthropic ↗
- We’re expanding the Claude Startups program to help founders build | Claude by Anthropic ↗
- Atlassian and OpenAI expand partnership to turn enterprise knowledge into action | OpenAI ↗
- Claude Frontier Academy: $100M to train 10,000 engineers \ Anthropic ↗
- Barclays scales Claude to upgrade operations and improve client experience \ Anthropic ↗
- October 2026 announcements - Partner Center announcements | Microsoft Learn ↗
- How Albertsons Companies is reimagining retail from the inside out | OpenAI ↗