Skip to content
NewsProduct launch

GitHub Copilot and Microsoft 365 Copilot add Claude Opus 5 for agentic coding tasks

· by Pondero Newsdesk

The short version

Anthropic's Claude Opus 5 landed in GitHub Copilot and Microsoft 365 Copilot on July 24, 2026, extending the Anthropic-Microsoft partnership and giving enterprise subscribers a model that leads FrontierBench v0.1 at a 43.5 percent resolution rate.

GitHub Copilot and Microsoft 365 Copilot add Claude Opus 5 for agentic coding tasks

GitHub and Microsoft both shipped Claude Opus 5 on July 24, 2026, the same day Anthropic released the model. Enterprise teams that run multi-step agentic coding workflows inside Copilot now have Anthropic's strongest coding model available as a selectable option.

What

GitHub's changelog confirmed Opus 5 for Copilot Pro+, Max, Business, and Enterprise subscribers via the model picker in Visual Studio Code, Visual Studio, Copilot CLI, the GitHub Copilot cloud agent, github.com, GitHub Mobile on iOS and Android, JetBrains IDEs, Xcode, and Eclipse. GitHub described the model as built for "complex, long-running coding tasks that require careful reasoning, effective tool use, and reliable execution across multiple steps." Business and Enterprise administrators must enable the Claude Opus 5 policy in Copilot settings before it appears for their users.

On the same day, Microsoft's announcement added Opus 5 to the model selector in Word, Excel, PowerPoint, Copilot Chat, Copilot Cowork, and Copilot Studio. Per Microsoft, Opus 5 "improves on Opus 4.8 across agentic coding, professional knowledge work, and long-horizon reasoning." Rollout varies by region and tenant configuration, and Microsoft directed users to check the Microsoft 365 Roadmap for availability timing.

Billing in GitHub Copilot runs at provider API list price under usage-based billing, per GitHub's pricing documentation. Microsoft did not disclose separate per-task pricing for Opus 5 in its announcement.

Why it matters

The benchmark gap is the clearest signal of what operators are getting. On FrontierBench v0.1, an agentic task benchmark, Opus 5 leads the leaderboard at 43.5 percent resolution rate. The second-placed GPT-5.6 Sol sits at 34.4 percent, and the previous Anthropic flagship, Opus 4.8, scored 21.1 percent. GitHub's early testing noted the model was "especially effective at making targeted changes, validating its work, and reducing unnecessary execution overhead on complex tasks."

For Copilot Business and Enterprise subscribers, the addition costs no new license. Tasks draw from usage-based credits, so the real decision is whether the per-task quality uplift justifies the cost premium over lighter models for a given workflow. The parallel rollout across GitHub Copilot and Microsoft 365 Copilot extends a partnership that previously brought Claude Fable 5 and Opus 4.8 into the same ecosystem.

Operators who want to evaluate Opus 5 inside the Copilot model picker can access it through GitHub Copilot.

What to watch next

Rollout is gradual, so some users may not see Opus 5 in the model picker immediately. The more substantive question is how Copilot's usage-based cost for Opus 5 compares on a per-task basis to running the same model directly through Anthropic's API via Claude Code or Cursor, particularly for high-volume agentic workloads where per-task cost accumulates quickly.

Sources