How is GitHub Copilot Fusion billed?

GitHub bills Copilot Fusion for each model it uses, at that model's standard rate, so a Cascade or Critique run can touch more than one model. For a limited time, users on paid Copilot plans get a 30% discount on model costs when they use Copilot Fusion.

What happened to HydraFusion and auto model selection in Copilot?

GitHub Copilot Fusion replaces both. People in the research preview who had enabled auto, HydraFusion or HydraFusion Max are moved automatically to the matching Copilot Fusion effort level: Standard, High or Max.

Can enterprise admins control which models Copilot Fusion uses?

Yes. Copilot Fusion only uses models enabled in enterprise and organization model policies. Enterprise admins can see which models Fusion chooses from, disable a model for Fusion without disabling it for everyone, and make Fusion at any effort level the default model through enterprise managed settings.

Can Copilot Fusion use local models?

Partly. An experimental, opt-in preview lets Copilot Fusion route between on-device and cloud models in Copilot CLI, the Copilot app and VS Code, using MAI Code 1.1 Flash through Windows ML or a model like Qwen3.8 Flash Next on a local OpenAI-compatible server. Fusion still makes its routing decisions in the cloud, so the session is not offline.

GitHub · 2026-10-28 · major

GitHub Copilot Fusion — Copilot now picks and combines models for you

GitHub Copilot Fusion is now generally available on all Copilot plans. Instead of picking a model, you pick Fusion and an effort level, and Copilot routes the task to one or more models. It grew out of the HydraFusion research preview.

GitHub changelog banner for the Copilot Fusion general availability post

GitHub Copilot Fusion replaces the model picker with routing: one model, a cheap-first cascade, or a draft-and-review loop.

Quick facts

Status	Generally available, 28 October 2026
Plans	All Copilot plans
Clients	VS Code, Copilot app, Copilot CLI, JetBrains IDEs
Effort levels	Fusion Standard, High, Max
Billing	Each model used at its standard rate; 30% off for a limited time on paid plans
Replaces	HydraFusion and auto model selection previews

What is it?

Copilot Fusion became generally available at GitHub Universe on 28 October 2026. It merges Project HydraFusion's multi-model orchestration with Copilot's fine-tuned auto model selection, so a user selects Fusion plus an effort level (Standard, High or Max) instead of a specific model.

How does it work?

Fusion looks at what a task needs across reasoning, code generation, debugging and tool use, then runs it in one of three patterns. Single sends it to one model. Cascade lets an efficient model draft and escalates to a stronger model if a quality check fails. Critique has one model draft, another review, and the first revise. An experimental opt-in preview can also route between on-device and cloud models.

Why does it matter?

GitHub says 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 API cost and tokens per task and about 40% less time. Enterprise admins keep control: Fusion only uses models allowed by policy.

Who is it for?

Copilot users and the admins who manage Copilot model policy

Frequently asked questions

How is GitHub Copilot Fusion billed?: GitHub bills Copilot Fusion for each model it uses, at that model's standard rate, so a Cascade or Critique run can touch more than one model. For a limited time, users on paid Copilot plans get a 30% discount on model costs when they use Copilot Fusion.
What happened to HydraFusion and auto model selection in Copilot?: GitHub Copilot Fusion replaces both. People in the research preview who had enabled auto, HydraFusion or HydraFusion Max are moved automatically to the matching Copilot Fusion effort level: Standard, High or Max.
Can enterprise admins control which models Copilot Fusion uses?: Yes. Copilot Fusion only uses models enabled in enterprise and organization model policies. Enterprise admins can see which models Fusion chooses from, disable a model for Fusion without disabling it for everyone, and make Fusion at any effort level the default model through enterprise managed settings.
Can Copilot Fusion use local models?: Partly. An experimental, opt-in preview lets Copilot Fusion route between on-device and cloud models in Copilot CLI, the Copilot app and VS Code, using MAI Code 1.1 Flash through Windows ML or a model like Qwen3.8 Flash Next on a local OpenAI-compatible server. Fusion still makes its routing decisions in the cloud, so the session is not offline.

Try it

Update VS Code, JetBrains, Copilot CLI or the Copilot app, open the model picker and select Copilot Fusion