Copilot Fusion Max reportedly comes within two points of Claude Code with Opus 5.5 Max at about half the API cost per task

On Oct. 28, 2026, GitHub reported that Copilot Fusion Max scored within two points of Claude Code with Opus 5.5 Max on Artificial Analysis’s Coding Agent Index at about half the API cost per task. In a reply, GitHub disclosed that it commissioned and funded the evaluation.

Artificial Analysis separately reported that GitHub Copilot Fusion used 7.2 million tokens per task on the index, approximately half the tokens used by Claude Opus 5.5 in Claude Code. Its token comparison names Copilot Fusion, not specifically Fusion Max.

GitHub also said users can choose High or Max effort and that Copilot Fusion can coordinate multiple models and workflows when one model alone is not enough. It advertised 30% off Copilot Fusion usage through Nov. 30.

The Coding Agent Index combines DeepSWE v1.1 (113 tasks), Terminal-Bench 4.0 (66) and SWE-Atlas-QnA (124), weighting the three benchmarks equally. Each component averages results across three attempts per task. Artificial Analysis defines cost per task as average pay-per-token API cost across the index, not consumer-plan pricing or the full cost of deploying a system in production.