AI TrackerTélécharger

Actus infrastructure,
pour les devs.

Les dernières actus « infrastructure » de tous les fournisseurs d'IA : lancements, changements de prix et retraits, résumés avec ce qu'ils changent pour vous.

OpenAI· Infrastructure

OpenAI explique comment l’entreprise a fait évoluer Habitat, sa plateforme de stockage, sur Azure Cosmos DB

Dans la deuxième partie d’une série consacrée à l’ingénierie, publiée le 23 octobre 2026, OpenAI présente Habitat, la plateforme de stockage en ligne qui sous-tend ChatGPT, l’API et Codex. OpenAI indique qu’Habitat sert plus d’un milliard d’utilisateurs dans 40 régions et stocke plus de 500 pétaoctets de données. Selon le résumé de l’article par OpenAI, les lectures de ChatGPT sont 12 fois plus rapides.Code

Polylane lance des agents proactifs qui surveillent et corrigent les applications Cloudflare Workers

Boris Tane a annoncé Polylane for Cloudflare, un service qui exécute des agents proactifs surveillant en continu les Workers, Durable Objects, Workflows et ressources similaires d’un compte Cloudflare. Selon l’annonce, les agents corrigent les bugs, améliorent les performances et gèrent les alertes.Infrastructure

Vercel explique comment Vercel Agent accède à la production

Vercel a publié un article sur la façon dont Vercel Agent accède aux déploiements en production. Selon Vercel, l’agent enquête sur un problème avec un accès en lecture seule, puis propose un plan ainsi que les autorisations nécessaires pour appliquer la correction. Ainsi, lui demander de corriger un déploiement ne revient pas à lui confier des identifiants à privilèges étendus.Infrastructure

Vercel AI Gateway ajoute tako_search, un outil de recherche de données structurées

Vercel a annoncé que la recherche de données structurées de Tako est désormais disponible comme outil sur AI Gateway. Au lieu de pages de texte, l’outil renvoie les données au format JSON, avec des unités, des sources et des dates : Vercel cite les actions, les sondages, la météo et les contrats publics, et précise que les données sont sous licence, sourcées et à jour.Anthropic· Infrastructure

Claude Models sur Google Cloud (géré par Anthropic) est désormais disponible de manière générale

Anthropic annonce que Claude Models sur Google Cloud (géré par Anthropic) est désormais disponible de manière générale. Selon Anthropic, cette offre apporte l’ensemble des fonctionnalités de Claude Platform sur Google Cloud, avec les identifiants IAM, la facturation et l’imputation des engagements de Google Cloud.Infrastructure

Vercel Drop déploie un fichier, un dossier ou un ZIP sans compte

Vercel a annoncé Vercel Drop : l’utilisateur glisse-dépose un fichier, un dossier ou un ZIP et obtient une URL publique, sans avoir besoin de compte. Le projet reste en ligne uniquement si l’utilisateur se connecte pour le revendiquer dans l’heure.Infrastructure

You.com lance Inference Router, une API compatible avec OpenAI pour les modèles à poids ouverts, avec recherche web intégrée

Le PDG de You.com, Richard Socher, a annoncé Inference Router, un service d’inférence hébergé doté d’une API compatible avec OpenAI et sans frais de plateforme. Le code existant compatible avec OpenAI peut être adapté en changeant l’URL de base.Google· Infrastructure

Google Cloud announces Agent Compute Engine and expanded agent governance for Gemini for business

Google Cloud announced a new product, Agent Compute Engine, at AGNTCon + MCPCon. It also announced expanded agent governance for Gemini for business. Google Cloud says the two together give organizations the control they need to run, govern and scale their agents. The details are in Google Cloud's blog post, "Introducing Agent Compute Engine and expanding agent governance".Anthropic· Code

Anthropic releases Claude API version 2026-10-21, with breaking changes

Anthropic's ClaudeDevs account announced API version 2026-10-21, which it calls the first official Claude API version release since 2023. Anthropic says it plans to release versions more often, so developers can handle breaking changes before a new model requires them. The older version, 2023-06-01, is not deprecated.OpenAI· Infrastructure· 1 discussion

OpenAI will require approval for model routers and AI gateways that offer its models

OpenAI says it has spent the past few months aligning safety policies with the partners and customers that provide its models, whether they host them or resell access. Partners must maintain safeguards consistent with OpenAI's own, investigate potential violations, take action and coordinate enforcement with OpenAI.OpenAI· Infrastructure

OpenAI API lets organizations choose which service tiers projects can use

OpenAI's API changelog for October 21, 2026 says that Platform settings now let you choose which service tiers projects can use, and that organization settings can restrict service tiers for all projects. The entry points to documentation on service tier errors.OpenAI· Infrastructure

OpenAI describes how ChatGPT's ad stack works

OpenAI published a post, summarized by @OpenAIDevs on X on October 21, on how ChatGPT's advertising system is built. OpenAI says it is designed to respond to the evolving conversation while keeping answers independent of ads.Infrastructure

Vercel makes WebSocket support in Vercel Functions generally available

Vercel announced that WebSocket support in Vercel Functions is now generally available. Vercel says it now serves 15 million WebSocket connections a day across AI chat, streaming and realtime multiplayer apps. The details are in Vercel's changelog entry.AlerteOpenAI· Modèles· 5 discussions

OpenAI says its models helped find critical security flaws in PHP

OpenAI announced that its models helped researchers, working under its "Patch the Planet" initiative, discover critical security flaws in PHP, which it says runs an estimated 70% of websites. According to OpenAI, it led a coalition of maintainers, researchers and security providers for about two months to prepare fixes and protections before disclosure; the fix was intricate and required a major change to the PHP engine.OpenAI· Modèles· 4 discussions

OpenAI says its models are coming to Google Cloud

OpenAI has published a post titled "OpenAI models are coming to Google Cloud". Its summary says OpenAI and Google Cloud are bringing frontier models and Codex workflows to Gemini's Model Garden, available through Google Cloud Marketplace. This is OpenAI's own account. Posts on X about the announcement say the models and Codex arrive on Google Cloud in the coming weeks.DeepSeek· Modèles

Stork: Alibaba raised its price for DeepSeek V4 Flash 0731 by 230%

Stork AI Daily, a newsletter that runs its own LLM price catalog, reports that Alibaba raised the price of DeepSeek-V4-Flash-0731 overnight. This is Stork's account from its own tracker; neither Alibaba nor DeepSeek has announced it, and the newsletter does not say which Alibaba service it means.Code· 1 discussion

Vercel introduces durable-harness for running coding-agent harnesses durably

Vercel announced durable-harness on X. In Vercel's words, it lets users run Pi, OpenCode, Codex, Claude Code, eve, fx or their own harness durably on Vercel, and agents automatically resume work after failures and run for hours, days or months.OpenAI· Infrastructure

OpenAI API adds a 12-hour prompt caching option

OpenAI's API changelog for October 20, 2026 says developers can now choose 12-hour prompt caching in the API (labelled for v1/responses). It keeps shared instructions, tools and context cached longer, so cached-input pricing applies even when requests are hours apart. OpenAI says 12-hour cache writes cost 2.5x the input token price.DeepSeek· Infrastructure

DeepSeek open-sources DeepInfer, its inference engine

DeepSeek has published DeepInfer on GitHub (deepseek-ai/DeepInfer) under the Apache-2.0 licence. The README calls it the inference framework that powers DeepSeek's large-scale production services. The release was shared on Hacker News on October 20. The details below are from the repository README.GitHub Copilot· Infrastructure

GitHub adds a REST API for Copilot model policies

GitHub's changelog announces generally available GET and PATCH REST endpoints to read and update Copilot model policies at four levels: enterprise, enterprise teams, enterprise apps and organizations (for example /orgs/{org}/copilot/model-policies).
Actus plus anciennes