GitHub Launches Copilot Fusion, an Orchestration Layer Between Anthropic and OpenAI Models

GitHub has released Copilot Fusion in GitHub Copilot, a runtime orchestration system that selects and coordinates models to optimize performance, cost, and latency for each task. The system chooses between a number of models from Anthropic and OpenAI, including Claude Opus 5.5, Claude Sonnet 5.5, GPT-6 Astra, and GPT-6 Sol. Token costs are charged based on the underlying models used to complete a request. Artificial Analysis benchmarked the max setting in GitHub Copilot CLI in its Coding Agent Index v2.0. Copilot Fusion scores 2 points below Opus 5.5 in Claude Code at roughly half the cost per task. It costs $5.59 per task, versus $11.42 for Claude Code with Opus 5.5 (max), while Devin Fusion with Opus 5.5 (max) + GPT-6 Luna (high) scores 62 at $8.00 per task. On the individual benchmarks, Copilot Fusion leads Terminal-Bench 4.0 with a score of 71. It ranks third on FrontierCode at 51, behind Claude Code with Opus 5.5 (max) at 56 and Devin Fusion at 54, while its SWE-Atlas-QnA-AA score lags at 57. Copilot Fusion also used 7.2M tokens per task, approximately half of the tokens used by Claude Opus 5.5 in Claude Code and 30% less than Devin Fusion. For developers choosing a coding agent, the results position Copilot Fusion as a cost-efficient option delivering near-frontier performance.