GitHub Copilot adds Claude Opus 5 for long-running coding agents

GitHub is gradually adding Anthropic’s Claude Opus 5 to Copilot development environments and its cloud agent for complex, long-running coding work.

07/25/20261 sources reviewed

Quick summary

The gradual rollout covers VS Code, Visual Studio, JetBrains, Xcode, CLI, and Copilot cloud agent, among other surfaces.
It is available to Pro+, Max, Business, and Enterprise plans; organization administrators must enable its policy.
GitHub reports strengths in agentic coding, while usage-based pricing and additional limits on some cyber requests apply.

WHAT HAPPENED

What happened?

GitHub is gradually adding Anthropic’s Claude Opus 5 to Copilot development environments and its cloud agent for complex, long-running coding work.

A larger model menu does not replace repository-specific tests, least-privilege access, or human review. GitHub’s early evaluation is a vendor result, so teams should compare cost, latency, and regressions on their own codebases.

WHY IT MATTERS

Why does it matter?

Developers can compare a new model focused on long-horizon tasks and tool use without leaving the Copilot workflow.

WHO SHOULD CARE

Who should care?

Software developersEngineering leadsEnterprise administrators

RELATED AI

Related AI

AIZIGOO VIEW

AIZIGOO view

Availability is gradual; check GitHub’s current pricing and supported surfaces.