GitHub Copilot expands model choice and content protections in this weekly release. VS Code adds new ways to manage agent sessions and get pull requests merge-ready. The release also includes general Claude Fable integration.

GPT-6 Astra from OpenAI is now generally available in GitHub Copilot. This is OpenAI's latest general-purpose model designed for long-horizon, autonomous coding and agentic tasks.

Gemini 3.8 Flash, Google's latest Flash model, is now available in GitHub Copilot. In early testing, Gemini 3.8 Flash performed strongly on complex terminal-based coding tasks and demonstrated rigorous capabilities.

Content exclusions are now generally available in the GitHub Copilot app and Copilot CLI. The feature respects content exclusion policies configured by enterprise, organization, and repository administrators. Copilot will no longer use excluded files as context, helping protect sensitive information.

Claude Fable 5.1 from Anthropic is now generally available in GitHub Copilot. It is the latest model in Anthropic's Mythos class, designed for long-horizon, autonomous coding and knowledge-work tasks.

GitHub Copilot in VS Code received updates across VS Code versions v1.132 through v1.135 throughout August 2026. These releases improve the ability to organize agent sessions, review changes, and navigate long conversations. Agent Host was also mentioned as part of the updates.

GitHub Copilot received weekly updates giving users more control over how Copilot runs. New features include team sessions in Slack and Teams, as well as customization options across the app, CLI, and IDE.

The August 2026 update for GitHub Copilot in Visual Studio introduces more control over how Copilot reasons, which AI models users can select, how teams share specialized agents, and when to request a code review.

The Customize tab in the GitHub Copilot app is now generally available. It brings MCP (Model Context Protocol) integration, allowing Copilot to work with the tools, knowledge, and workflows teams already rely on. This update enhances Copilot's ability to connect with external tools and services.

GitHub Copilot's agentic capabilities are now available in Slack through the GitHub integration, entering public preview. Users can interact with @GitHub in Slack to access GitHub Copilot CLI and GitHub Copilot app features directly within their Slack workspace.

Grok 4.6, xAI's latest reasoning model, is now rolling out in GitHub Copilot. It is designed for agentic coding and complex multi-step workflows.

GitHub Copilot received weekly updates including new models and portable plugins. The release also brings smoother agent workflows and improved flexibility across editors, the command line, and the Copilot app.

Gemini 3.7 Flash, Google's latest Flash model, is now rolling out in GitHub Copilot. Early testing shows improvements in web and app development and agentic capabilities.

This update introduces persistent memory and local model access via Ollama to GitHub Copilot for JetBrains. It also adds more enterprise controls and improves everyday chat workflows. Additionally, reliability issues across MCP servers have been resolved.

GitHub Enterprise Server 3.22 release candidate introduces new capabilities across the platform. One notable highlight is that administrators can now configure Copilot CLI as part of the new release features.

GitHub Copilot is launching MAI-Code-1.1-Flash as a replacement for MAI-Code-1-Flash. The older MAI-Code-1-Flash model will be deprecated across all GitHub Copilot experiences on September 10, 2026. Users are advised to update their workflows to use the new MAI-Code-1.1-Flash model.

GitHub Copilot received weekly updates across the desktop app, CLI, and VS Code. The updates focus on helping users resume and organize work, review changes, and ask questions without losing context.

Lite and Balanced effort levels for GitHub Copilot code review are now generally available. These effort levels allow users to match the depth of a code review to the complexity and risk of the changes being reviewed.

Kimi K3 model has been made available in GitHub Copilot after resuming its rollout. This model is billed at provider list pricing under usage-based billing.

Enterprise owners can now centrally control which Model Context Protocol (MCP) servers GitHub Copilot clients are allowed to run. This is managed through new allowedMcpServers and deniedMcpServers keys in enterprise managed settings. This provides organizations greater security and governance over MCP server usage within Copilot clients.

GitHub now offers automatic code coverage workflow generation using AI directly from a repository's Code Quality settings. This removes the need for manual configuration, allowing developers to enable coverage tracking faster and with less setup effort. The feature is surfaced as a new option in the repository settings UI.

GitHub Copilot received multiple updates across VS Code versions 1.127 through 1.131 throughout July 2026. The releases improve how users work with agents, review changes, use chat, and navigate VS Code.

GitHub Copilot in Visual Studio received a July 2026 update featuring a new agent built on the Copilot SDK. The update includes built-in expertise from the .NET and Azure teams, along with more ways to tailor GitHub Copilot to individual and team workflows.

Copilot code review support for agent skills and MCP (Model Context Protocol) servers is now generally available for all Copilot Pro, Pro+, Business, and Enterprise users. This feature was previously announced in public preview. The update expands access to these capabilities across all Copilot subscription tiers.

Grok 4.5, xAI's latest reasoning model, is now available in GitHub Copilot. It is designed for fast, agentic coding and complex multi-step workflows. The model features a large context window for handling extensive coding tasks.

GitHub Copilot for JetBrains now supports connecting MCP servers and custom agents in Claude agent flows. The update also introduces improved OpenTelemetry configuration for better telemetry tuning and enhanced model management capabilities.

Claude Opus 5, Anthropic's newest Opus model, is now available in GitHub Copilot. It is designed for complex, long-running coding tasks that require careful reasoning, effective tool use, and reliable performance.

The GitHub MCP Server now supports the next MCP specification ahead of its official release on July 28, 2026. The MCP protocol is transitioning to a stateless core architecture, and GitHub has implemented this new spec early. This affects how GitHub Copilot and other tools interact with the MCP server for tool use and integrations.

Copilot cloud agent for Linear is now generally available. Users can assign Linear issues directly to Copilot's asynchronous, autonomous background agent. When assigned an issue, Copilot will analyze the issue contents and work on it autonomously.

Gemini 3.6 Flash, Google's latest Flash model, is now rolling out in GitHub Copilot. It is designed for web and app development, coding, and longer-horizon agentic tasks. It has configurable settings for these use cases.

GitHub Code Quality has reached general availability on GitHub Enterprise Cloud and GitHub Team plans. The feature is explicitly positioned as a complement to AI-accelerated code output (including Copilot), helping teams maintain code quality as AI increases code generation volume. This represents a platform-level capability expansion directly tied to managing AI-generated code at scale.

The June 2026 update for GitHub Copilot in Visual Studio focuses on visibility and trust, introducing a clearer view of Copilot usage and a new trust layer for MCP servers. It also includes the first C++ scenarios for Copilot in Visual Studio.

GitHub Copilot now includes a /security-review slash command available in public preview within the GitHub Copilot app. This feature allows users to run security reviews on in-flight code changes directly from the app. It brings the same security review capabilities previously available elsewhere into the Copilot app interface.

GitHub Copilot for JetBrains IDEs received major advances in customization and model provider flexibility across all tiers. The update brings richer plugin and provider experiences, improved conversational interactions, and expanded BYOK (Bring Your Own Key) capabilities.

GitHub Mobile now includes improved filters and sorting for Copilot sessions, making it easier to find the right session as your session list grows. Users can now narrow down their sessions more efficiently as their Copilot session list expands.

OpenAI's GPT-5.6 family is now rolling out in GitHub Copilot. GPT-5.6 comes in three variants: Sol, Terra, and Luna, allowing users to match the model to the specific job at hand.

JetBrains and GitHub announced a deeper integration between JetBrains AI Assistant and GitHub Copilot. Copilot Agent is now available directly within JetBrains AI Assistant, expanding access for millions of developers who use JetBrains IDEs.

Claude Sonnet 5, Anthropic's latest Sonnet-class model, is now generally available in GitHub Copilot. It brings strong coding performance to everyday development and agentic workflows. This gives developers a new Sonnet-class model option within Copilot.

MAI-Code-1-Flash, Microsoft AI's in-house coding model, is now generally available for GitHub Copilot Business and Copilot Enterprise. It is purpose-built for coding and optimized for performance, building on its recent expansion across Copilot surfaces.

GitHub Copilot is now integrated into GitHub Desktop 3.6, powering commit authoring and merge conflict resolution. This brings more of the day-to-day Git flow into one place, alongside new Git worktree support.

Copilot code review now uses built-in file exploration tools available in the Copilot CLI and SDK. This update significantly improves review cost efficiency with no change to existing user workflows.

The GitHub Copilot app now supports bring your own key (BYOK), allowing users to run agent sessions against their own model providers. Supported providers include OpenAI, Azure OpenAI, Microsoft Foundry, and Anthropic, among others. This gives users more flexibility and control over which models power their Copilot agent sessions.

The redesigned terminal interface for GitHub Copilot CLI, previously previewed at Microsoft Build 2026, is now generally available. It features a tabbed layout for working with GitHub directly from the terminal.

This update adds support for organization and enterprise agents from GitHub in JetBrains IDEs. It introduces the ability to queue and steer messages in Copilot CLI sessions, and adds a new agent debug logs summary view. Claude is also being introduced as an agent provider in preview.

Copilot code review received new configurations and controls including organization runner controls, Copilot content exclusion support, and removal of the character limit on repository custom instructions. These updates make it easier to tailor Copilot code review to specific organizational needs.

GitHub Copilot CLI introduced a new /settings slash command that provides a unified, schema-driven configuration home. This update consolidates previously scattered commands like /theme, /streamer-mode, and /experimental into a single place, simplifying the configuration experience for users.

GitHub Agentic Workflows is now in public preview, enabling automation of reasoning-based tasks using coding agents. The feature supports tasks like issue triage, CI failure analysis, and documentation updates. This expands GitHub Copilot's agentic capabilities to a broader audience through public preview availability.

GitHub improved the handoff experience between Copilot Chat and Copilot cloud agent on the web. New functionality was enabled allowing users to search and query past agent sessions, improving continuity and accessibility of previous work.

Claude Fable 5 from Anthropic is now generally available in GitHub Copilot. It is the first model in Anthropic's Mythos class, designed for long-horizon, autonomous coding and knowledge-work tasks.

GitHub Copilot in Visual Studio received its May 2026 update, bringing sharper tooling for planning, reviewing, and managing work. The update is described as having highlights for all GitHub Copilot users in Visual Studio 2026.

Copilot Pro, Pro+, and Max users can now programmatically start and track Copilot cloud agent tasks using the new Agent Tasks REST API, now available in public preview. This API enables developers to interact with Copilot cloud agent functionality through REST endpoints.

GitHub Copilot in Visual Studio Code shipped weekly stable releases v1.120 through v1.123 throughout May and early June 2026. The releases included updates making Agents features more prominent. The most recent release in this series is v1.123.

GitHub Copilot now introduces Copilot CLI availability in JetBrains IDEs, bringing new capabilities for Copilot CLI sessions. The update also delivers broader agentic improvements to enhance the overall Copilot experience within JetBrains IDEs.

GitHub Copilot code review is now available in technical preview for Azure Repos, bringing on-demand pull request reviews directly into Azure DevOps workflows. This extends Copilot's code review capabilities beyond GitHub.com to Azure DevOps users.

The GitHub Copilot app technical preview has been expanded to all existing Copilot Pro, Pro+, Business, and Enterprise customers. The desktop app is now available for download on Windows, macOS, and Linux platforms. This represents a significant expansion of availability from a limited preview to the full customer base.

GitHub Copilot CLI received a major refresh announced at Microsoft Build 2026. New generally available features include rubber duck mode, prompt scheduling, and voice input. An experimental new terminal interface with tabs was also introduced.

The GitHub Copilot SDK is now generally available (GA). Developers can embed GitHub Copilot's agentic engine into their own applications, services, and developer tools. The release comes with a stable API and production-ready support.

Claude Opus 4.8, Anthropic's latest Opus model, is now generally available in GitHub Copilot. Early testing demonstrates a clear step forward in code understanding and generation.

Copilot Memory now includes improved memory deletion capabilities and adds a repository-level off switch for more granular control. Additionally, further memory controls have been brought into the Copilot CLI, with the feature remaining in public preview.

GitHub Copilot Chat on web now supports natural language semantic issue search, allowing users to quickly find, group, and analyze issues. The feature is powered by a new semantic issues index that provides context-aware results.

GitHub updated the available model selection for Copilot Chat on the web to deliver more consistent, high-quality responses. The change involves limiting model choices to improve response quality and consistency.

Gemini 3.5 Flash, Google's latest Flash-tier model, is now generally available on GitHub Copilot. It delivers near-Pro coding quality at Flash-tier speed and cost, making it a significant addition to Copilot's available models.

GitHub introduced a new REST API endpoint in public preview that allows programmatic auditing of a repository's Copilot cloud agent configuration. This 'Get Copilot cloud agent configuration for a repository' API enables developers and administrators to retrieve configuration details via REST calls.

The GitHub Copilot Spaces API is now generally available. Developers can now programmatically create, read, update, and delete Spaces directly from their own applications. This moves the feature from preview to full general availability.

GPT-5.3-Codex is now the base model for all Copilot Business and Copilot Enterprise organizations, replacing GPT-4.1. This change affects the default model used when no specific model is selected by the organization.

The GitHub Copilot app is now available in technical preview, offering a GitHub-native desktop experience for agentic development. It allows users to start agentic development from the work in front of them and keep it isolated. The app enables users to steer agentic workflows directly from the desktop.

Copilot cloud agent now supports auto model selection. When users select 'Auto' in the model picker, Copilot intelligently selects the best available model based on system health and model performance criteria.

Copilot Business and Copilot Enterprise users can now programmatically start Copilot cloud agent tasks using the new Agent Tasks REST API, which is available in public preview. This update enables automated and programmatic access to Copilot cloud agent capabilities.

This update introduces the Copilot CLI agent into JetBrains IDEs, bringing CLI-based agentic capabilities to the JetBrains ecosystem. It also adds a unified sessions view that displays live status for running and queued sessions. Additional features were mentioned but not fully detailed in the excerpt.

Autonomous agent mode. Plan, build and test features from a GitHub issue with minimal input.

Conversational coding assistant added. Ask questions about your code directly in the editor.

First AI pair programmer. Trained on public GitHub repos using OpenAI Codex. Technical preview launched.