From the source
GitHub made Copilot Fusion generally available, letting Copilot route and orchestrate which models handle each task instead of users picking one.
- The system combines Project HydraFusion's multi-model orchestration with Copilot auto model selection's fine-tuned model selection, which GitHub says balances performance and cost efficiency.
- GitHub says Copilot Fusion Max scored 60 on the Artificial Analysis Coding Agent Index, within two points of Claude Code with Opus 5.5 Max, at about half the average API cost per task, using about half as many tokens and about 40% less time per task.
- Users select Copilot Fusion in the model picker and then choose an effort level: Fusion Standard for everyday coding, Fusion High for challenging tasks and deeper reasoning, and Fusion Max for the most demanding tasks.
- Copilot Fusion can solve a task with one model, or use Cascade, where an efficient model drafts and a quality check decides whether to escalate, or Critique, where one model drafts, another reviews, and the first revises.
- Administrators control which models Copilot Fusion may use and can set it as the default through Enterprise managed settings, and it only uses models enabled in enterprise and organization model policies.
