GitHub Copilot Fusion routes coding work between Anthropic and OpenAI models
GitHub Copilot Fusion switches coding tasks between Anthropic and OpenAI models mid-task instead of fixing developers to one model.
GitHub Copilot Fusion routes coding work between models from Anthropic and OpenAI, so developers don't have to pick one model for a task. According to the reporting, it switches coding tasks between the two vendors' models mid-task instead of locking a developer to a single model. Artificial Analysis says Fusion's top configuration scored within two points of Claude Code Opus 5.5 Max while costing roughly half as much per task.
The available details don't say which benchmark produced the two-point gap or what the cost-per-task figure covers. They also don't say how Fusion decides when to switch models. Those details matter, since a score gap and a cost ratio only mean something against the task set and pricing behind them.
What changed
GitHub Copilot Fusion now routes coding tasks between Anthropic and OpenAI models mid-task instead of fixing developers to one model.
Type: New capability · Applies to: GitHub Copilot Fusion
Why it matters
Teams standardizing on one coding model may need to test multi-model routing, since the reported cost per task is about half of a single-model setup.
What you can do
Benchmark multi-model routing against your current coding setup on your own tasks before changing defaults.
Original source
We have no link to the original source for this story yet.
Where we saw it
Forward Future (Matthew Berman)
AI wrote this summary from the sources listed, and nobody reviewed it individually before publication. Check the original source before you rely on it. How it works