The latest AI news and updates of GitHub — Developer platform powering GitHub Copilot and AI-assisted software development workflows. Covering GitHub's latest product updates, company news, and analysis from the past 90 days.
GitHubSep 9GitHub Copilot code review now submits formal pull request approvals, addressing the 84% of PR lifecycle time spent waiting for sign-off. Administrators can enable this feature at the enterprise, organization, or repository level, with approvals counting toward required-approvals rules. Copilot automatically dismisses approvals if new commits are pushed, and the feature is currently in public preview.
GitHubSep 8GitHub updated its Copilot agent harness to reduce token waste without sacrificing task quality. By selectively compressing output, removing redundant line-number formatting, and batching background notifications, the system avoids unnecessary model turns. A meta-prompting loop also reduced prompt tokens by 1.8% per session. These changes prioritize task completion over minimizing individual tool-call tokens.
GitHubSep 6GitHub introduced canvases in the Copilot app to provide a durable, shared surface for agentic workflows. Canvases replace long chat scrolls by tracking what ran, changed, and requires human review. The feature includes example templates like Java Modernization Studio and Site Studio, available in awesome-copilot, to maintain state and governance during complex, multi-step AI tasks.
GitHubSep 4GitHub introduced Project HydraFusion, a research preview that dynamically orchestrates multiple AI models to execute coding tasks. In offline evaluations on Terminal-Bench 2.1, the system improved verified task quality by 4.9 percentage points while reducing estimated costs by 67% compared to Claude Opus 5. The preview is available through the /experimental flag in the GitHub Copilot CLI.
GitHubSep 3GitHub added Google's Gemini 3.8 Flash to Copilot, now available in the app, CLI, and VS Code. The model shows strong results on complex terminal-based coding tasks. Access extends to Pro, Pro+, Max, Business, and Enterprise plans with introductory provider pricing through December 31, 2026.
GitHubSep 2GitHub added Anthropic's Claude Fable 5.1 to GitHub Copilot for Pro+, Max, Business, and Enterprise plans. The model performs on long-running coding tasks, deep codebase research, and complex agentic workflows. It requires data retention by default for safety classifiers, though eligible enterprises can access zero-data-retention endpoints under a time-bound exception.
GitHubAug 22GitHub is offering a 30% discount on auto model selection for Copilot Max users. This feature automatically routes tasks to Claude, GPT, or Microsoft AI models without requiring manual code or opt-in. The promotional pricing remains active until 00:00 UTC on September 30, 2026.
GitHubAug 16GitHub released Lite and Balanced effort levels for Copilot code review, replacing the previous Low and Medium settings. Balanced provides deeper analysis for complex pull requests, while Lite targets straightforward changes. Organization-level defaults govern review depth across repositories, with per-request selection available for Pro, Pro+, Max, Business, and Enterprise plans.
GitHubAug 13GitHub is rolling out Google's Gemini 3.7 Flash model to GitHub Copilot. The model improves web and app development, agentic coding workflows, and codebase research. It is available to Copilot Pro, Pro+, Max, Business, and Enterprise users across IDEs including VS Code, JetBrains, and Xcode. Business and Enterprise administrators must enable the model policy to grant access.
GitHubAug 11GitHub is rolling out Microsoft's MAI-Code-1.1-Flash model to Copilot, featuring native vision support for image understanding and improved coding performance. The update delivers a 73% lower list price compared to the previous version. The model is available across Copilot interfaces, with manual selection options for Pro and Enterprise plans and auto-selection for Free and Student users.
GitHubAug 6GitHub added Moonshot AI’s Kimi K3, an open-weight model, to GitHub Copilot. The model features frontier-level agentic coding capabilities and is hosted on Fireworks AI infrastructure. It is rolling out to all paid plans, though Business and Enterprise administrators must manually enable the model policy. Pricing follows provider list rates, starting at $3 per million input tokens.
GitHubAug 5GitHub added voice prompting to the GitHub Copilot desktop app, replacing manual typing with spoken input. A new microphone icon in the app interface captures and transcribes speech directly into the session. This feature is available on macOS, Windows, and Linux for all Copilot plans.
GitHubAug 3GitHub’s Aspire team implemented GitHub Agentic Workflows to automate cross-repo documentation. The pipeline generates draft pull requests from merged features, requiring subject-matter expert review. The team merged 82 documentation pull requests with a 44.8-hour median time-to-merge, keeping documentation synchronized with code changes through automated intent-based processing.
GitHubJul 27GitHub updated Copilot with prompt caching and deferred tool loading to cut token overhead. HyDRA, its new routing architecture, selects models by task intent and real-time system health. The approach matched a 70.8% resolution rate at 3.3× cost savings — now live in VS Code, GitHub.com, and mobile.
GitHubJul 24GitHub added Anthropic's Claude Opus 5 to Copilot for Pro+, Max, Business, and Enterprise plans. Early testing shows the model performs well on agentic coding tasks, including autonomous code changes and regression verification, while reducing execution overhead on complex workflows. Business and Enterprise administrators must enable the model policy in Copilot settings to grant access.
GitHubJul 23GitHub argues that while AI makes writing code cheaper, the cost of owning it remains high. The company suggests replacing abstract scope debates with evidence from a constrained, AI-generated first patch. This method reveals risk and ownership costs in thirty minutes, shifting scope discipline from pre-implementation planning to concrete artifact review.
GitHubJul 17GitHub offers promotional pricing for GPT-5.5 within Copilot from 00:00 UTC July 18 until 00:00 UTC July 20. Copilot Max users receive a 60% discount, while Pro+ users receive 40% off. The discount applies automatically to all model usage during this period with no opt-in required.
GitHubJul 15GitHub Copilot for Jira reaches general availability, streaming agent progress directly into Jira tickets in real time. The integration adds post-session steering, where follow-up instructions in the Jira chat panel refine existing draft pull requests. A simplified setup process reduces the configuration steps required to connect GitHub organizations and repositories to the Jira application.
GitHubJul 14GitHub Mobile now supports live notifications for remote Copilot CLI sessions. The update provides real-time tracking of agent progress, alerts for required input, and direct access to session logs from mobile devices. These features facilitate remote management of terminal-based agent tasks while away from a desktop terminal.
GitHubJul 13GitHub updated Copilot CLI to version 1.0.42, implementing smarter subagent delegation that reduces tool failures by 23%. This update cuts search failures by 27% and edit failures by 18% while lowering P95 wait times by 5%. These changes improve responsiveness and reliability during multi-step coding tasks with no regression in output quality.
GitHubJul 11GitHub published the Multilingual Repositories Dataset, a metadata index covering over 40 million repositories and 80 million classification rows. The dataset identifies non-English READMEs, issues, and pull requests, with Korean leading in issue text and Portuguese in READMEs. This open-source resource provides researchers and developers with signals to study multilingual developer communities and build more inclusive AI tools.
GitHubJul 5GitHub Copilot CLI now supports custom agents defined in Markdown files. These profiles encode team-specific roles, tools, and guardrails into reusable, versioned assets. This update enables consistent, reviewable workflows for tasks like security audits, release notes, and incident response directly from the terminal.
GitHubJul 2GitHub added Moonshot AI’s Kimi K2.7 Code to the Copilot model picker, marking the first time an open-weight model is available for selection. The model offers a lower-cost alternative with performance comparable to frontier models. It is rolling out gradually to Pro, Pro+, and Max plans, with enterprise access requiring administrator policy enablement.
GitHubJul 2GitHub launched a Medium-depth analysis tier for Copilot code review, now in public preview. This tier routes complex pull requests to a higher-reasoning model for deeper analysis of logic and security-sensitive code. Repository admins configure the review effort level in settings, with the Medium tier consuming more AI Credits than the fast, cost-efficient Low tier.
GitHubJun 29GitHub benchmarked its Copilot agentic harness against native model-vendor harnesses across five benchmarks, including SWE-bench Verified and TerminalBench. The results show task resolution on par with native harnesses while consuming fewer tokens in most configurations. Copilot’s multi-model architecture supports over 20 frontier models, allowing for task-specific optimization between efficiency and peak quality.
GitHubJun 28GitHub Copilot code review now reads repository-level AGENTS.md files to generate context-aware feedback based on project conventions. The update also adds a Request button to draft pull requests for direct review access and collapses Copilot review events on the pull request timeline to declutter the conversation view. All features are generally available.
GitHubJun 25GitHub released an LSP Setup skill for Copilot CLI that automates the configuration of Language Server Protocol servers across 14 languages. The skill replaces brute-force text-search heuristics with structured semantic analysis, providing the agent with precise definitions, references, and type resolution. It supports both user-level and repository-level configurations to improve code accuracy during autonomous terminal tasks.
GitHubJun 23The GitHub Copilot app now supports Bring Your Own Key, integrating model providers like OpenAI, Anthropic, and local Ollama instances. The model picker assigns specific models to each agent session, with API keys stored in the local OS keychain. This update routes inference traffic through private cloud tenants and pairs frontier models with local execution for specific agent tasks.
GitHubJun 23GitHub Copilot CLI introduced a terminal user interface that enables direct management of GitHub issues, pull requests, and gists. The interface supports creating, reviewing, and merging changes directly within the terminal. This update integrates native GitHub platform capabilities into the command-line agent, supporting multi-step workflows and session persistence.
GitHubJun 22GitHub has integrated Impeccable, a design-and-quality tool with 40k+ GitHub stars, as a built-in skill in the Copilot app. The integration makes design and quality a native layer for all creators — no separate setup needed. AI raised the floor on coding speed; Impeccable targets the ceiling on polish and consistency.
GitHubJun 20GitHub integrated LLM-based contextual verification into its secret scanning pipeline to improve alert quality. By analyzing usage signals like variable assignments and API calls instead of full files, the system reduced customer-confirmed false positives by 75.76%. This focused approach filters out noise like test data and placeholders while maintaining detection performance.
GitHubJun 19GitHub Copilot CLI now includes generally available on-device speech-to-text models for voice interaction. The update also introduces a built-in Rubber Duck agent, which uses a secondary AI model to review plans, code, and tests for potential blind spots. These features operate directly within the terminal to support agentic development workflows.
GitHubJun 18GitHub is rolling out Microsoft’s MAI-Code-1-Flash model across eight additional Copilot surfaces, including the Copilot CLI, the new desktop app, and major IDEs. Designed specifically for Copilot, the model delivers best-in-class quality for its size and outperforms other small models in early testing. It is currently available to Free, Student, Pro, Pro+, and Max plan users.
GitHubJun 17GitHub released the standalone Copilot desktop app for macOS, Windows, and Linux, centralizing agent-driven development from issue to pull request. The app introduces Canvases for steerable agent progress, cloud automations for scheduled agent work, and support for custom models and MCP servers. Enterprise access requires administrators to enable the Copilot CLI in policy settings.
Frequently asked questions