MiniMax retires its M2.5 and M2.7 models today · Azeem Hassni
AI-generated summary MiniMax retires its M2.5 and M2.7 model families today. Pay-as-you-go API requests to the retired models stop working today.
Search tips
"exact phrase" the words together, in this order-word leave out stories with the wordtopic:robotics a topic by ID or name; quote names with spaces: topic:"open-weight models"category:"AI agents" a categorycompany:openai a company, person, product or organization (also entity:)source:forbes.com a publisher by site or nametype:analysis news, analysis, opinion, research paper and other kindsafter:2026-11-01 stories on or after that New York day; before:2026-11-10 on or before itAI-generated summary MiniMax retires its M2.5 and M2.7 model families today. Pay-as-you-go API requests to the retired models stop working today.
AI-generated summary GitHub released browser use for Copilot in public preview on November 17, providing the agent with a live Chrome tab to click, type, and read. The capability functions across the Copilot app, Copilot CLI, VS Code, and JetBrains IDEs following the installation of GitHub's Copilot Chrome extension.
AI-generated summary GitHub has updated GitHub Copilot code review with three public preview features including an Adaptive effort setting and secret detection for over five hundred types. The company also moved its Azure DevOps integration to general availability and added review outcome reporting to Repository insights.
AI-generated summary Microsoft has introduced MAI Realtime 1 as the first speech-to-speech model from its Microsoft AI group, pricing it at $0.01 per audio minute for conversational use. Microsoft says MAI Realtime 1 outperforms OpenAI's GPT Realtime 2.1 across public benchmarks, and early testers include Agoda and Reservations.ai.
AI-generated summary GitHub explains that the Agent Host Protocol runs session continuity for GitHub Copilot across devices and applications. GitHub states that the protocol uses URI-identified channels and reducer-based state objects to manage interactions and reconnections.
AI-generated summary Microsoft announced the Microsoft Copilot SDK 2.0 at Ignite, providing programmatic access to the Microsoft Copilot harness. Microsoft says the new release includes a Rust core in-process runtime that hosts many sessions within a single process.
AI-generated summary OpenAI releases two new language models named GPT-6.2 Astra and GPT-6.2 Sol, featuring lower API prices and benchmark gains. OpenAI says GPT-6.2 Astra is designed for demanding professional tasks, while GPT-6.2 Sol is built for everyday use.
AI-generated summary Anthropic explains linked billing for Claude Platform accounts, which lets administrators connect a Claude Console organization to a third-party app without sharing an API key. Administrators can set monthly spend limits, view monthly costs, and unlink apps on the Linked platforms page in the Claude Console.
AI-generated summary Google shuts down Opal, its no-code tool for chaining text, image and video prompts into AI-powered mini-apps. Google states that existing workflows will not automatically migrate, meaning users must set up new skills or prompts manually.
AI-generated summary The vLLM 0.34.0 release enables default cache expansion to use actual free GPU memory and unifies GPU model execution paths for inference engine deployments. The vLLM project reports performance updates including throughput and latency changes across various hardware configurations such as H200 and GB200.
AI-generated summary MAI-Code-2 and MAI-Code-2-Flash appeared in GitHub Copilot's model catalog with 1.05M context according to reports. Microsoft updated MAI Playground with a redesign ahead of the Ignite keynote where new models are expected.
AI-generated summary GitHub shipped two changes on 16 November 2026 including a new managed setting that lets administrators require selected GitHub Copilot customizations to come from plugins or managed settings. VS Code 1.146 deprecated the Local harness which will be removed in February 2027 while the Copilot harness takes over to support managed sandbox and permission controls.
AI-generated summary Anthropic says it will reduce its coordinated vulnerability disclosure timeline for high- and critical-severity bugs to 30 days starting on 1 January 2027. Google Project Zero announced a plan to transition its vulnerability disclosure schedule down to 30 days by early 2028.
AI-generated summary Vercel released AI Gateway Private Inference on November 16, 2026, allowing teams to register custom model endpoints and call them through the gateway. Vercel states that private models work in coding agents connected to the gateway, such as Claude Code, Codex and Cursor.
AI-generated summary GitHub retired Claude Opus 4.8, Claude Opus 5 and GPT-5.6 Terra across most GitHub Copilot features on November 16. GitHub points to alternative models including Claude Opus 5.5, GPT-6.1 Sol and GPT-6 Luna.
AI-generated summary The FBI warned U.S. organisations that criminals are using Artex, an open source AI penetration-testing framework, to compromise financial firms, governments and law offices in under three hours. Microsoft says Storm-0940 used the framework to attempt sign-ins to more than 2,250 ConnectWise tenants, while Barracuda reported attackers used it to reach cloud accounts, an analytics system and Kubernetes clusters.
AI-generated summary Google plans to make custom Gems in Gemini read-only for adults using personal Google accounts on November 17 as it moves them to skills. Google says it will automatically move personal users custom Gems and eligible attachments to skills, where the saved instructions can be reused.
AI-generated summary OpenAI announced a new image classifier designed to detect child sexual abuse material and new warnings in ChatGPT to prevent users from accessing or creating harmful content. OpenAI says the classifier was developed with the National Center for Missing and Exploited Children and flags fewer false positives during testing.
AI-generated summary Hugging Face and OpenRouter state they are teaming up to make open-weight models easy to run. Hugging Face states it is switching Inference Providers from its own custom router to OpenRouter.
AI-generated summary Liquid AI released two omnimodal decision models named d1.2 and d1.2-ultra to handle text, audio, and image inputs. The top-tier d1.2-ultra model achieved the highest rank on the Artificial Analysis Decision Index v2.1 and the private AA-Paperclip benchmark.
The top AI news in your email delivered each morning. Choose categories