Updated 29 Oct 2026: Oct 28 Copilot Fusion GA, effort levels, model policy/billing and experimental local routing; correct earlier research-preview framing.

Key details

  1. October 28, 2026: Copilot Fusion generally available, replacing HydraFusion research-preview selection.
  2. Supported clients: VS Code, Copilot app, Copilot CLI and JetBrains.
  3. Three execution patterns remain Single, Cascade and Critique.
  4. Standard, High and Max effort levels control model selection and orchestration.
  5. Enterprise admins can restrict eligible Fusion models and choose defaults.
  6. Each model used is billed at its normal rate; limited-time 30% discount on model costs for paid plans.
  7. GitHub reports a Fusion Max Artificial Analysis score of 60; its cost and speed comparisons are not independently reproduced here.
  8. Local model routing is experimental; managed Windows ML and manually configured CLI options differ.
  9. Local inference does not make Fusion offline because routing is still cloud-based.

What builders should take away

  1. Test Fusion Standard/High/Max on representative tasks and compare total billed model calls, latency, review burden and accepted code.
  2. Verify enterprise model policies and spend budgets before enabling multi-model orchestration broadly.
  3. Do not assume the 30% model-cost discount applies indefinitely or makes every task cheaper.
  4. For local routing, check hardware requirements, supported models and client limitations; treat it as experimental.
  5. Keep cloud-data and prompt-handling requirements in place even when local inference is enabled.

What changed

On October 28, 2026, GitHub renamed and generally released its HydraFusion multi-model orchestration as Copilot Fusion across VS Code, Copilot CLI, Copilot app and JetBrains. It selects Single, Cascade or Critique execution per task and offers Standard, High and Max effort tiers. Enterprise administrators can restrict the eligible models and set a default. Billing charges each underlying model at its standard rate, with a limited-time 30% discount on model costs for paid plans. In a separate experimental preview, Copilot Fusion Standard can route to local models: managed Windows ML setup for MAI Code 1.1 Flash on supported Windows devices, or manually configured OpenAI-compatible local servers on Windows, macOS and Linux through Copilot CLI. Routing decisions still happen in the cloud, so local inference does not make sessions offline.

Why it matters

The original October 1 dossier described HydraFusion as a research preview for mainstream editors. General availability changes procurement and operational planning: teams can now select an effort policy instead of choosing a single model, restrict the model pool under enterprise governance and estimate costs from all participating model calls rather than a single model price. GitHub reports Fusion Max scored 60 on Artificial Analysis Coding Agent Index, within two points of Claude Code Opus 5.5 Max, at roughly half average API cost per task and 40% less time; these are GitHub-cited benchmark comparisons, not independently reproduced measurements of a team's workload. The local-model preview may reduce some cloud inference work but still requires cloud routing and suitable local hardware. Admins should verify which models, data policies and spend limits apply to their organisation.

October 28: research preview becomes Copilot Fusion GA

Fusion replaces the HydraFusion research-preview name and is available on all Copilot plans in VS Code, Copilot app, CLI and JetBrains. Standard favours versatile everyday models, High more capable models for harder tasks, and Max the strongest models for demanding work. The system can answer with one model, escalate from an efficient model after a quality gate, or have a read-only critic review and request a revision.

Costs and enterprise model controls are explicit

GitHub bills each underlying model at its normal rate, not a flat fee per Fusion turn; paid Copilot users temporarily receive 30% off model costs through Fusion, with no published end date in the announcement. Administrators can enable or disable particular models for Fusion independently of their general availability, and set Fusion as a default through managed settings. Compare total task cost, latency and review quality rather than just advertised model multipliers.

Experimental local routing is not an offline Copilot

On supported Windows devices, Windows ML can download and configure MAI Code 1.1 Flash. Copilot CLI also supports manually configured MAI Code 1.1 Flash or Qwen3.8 Flash Next through an OpenAI-compatible local server on Windows, macOS or Linux. The Copilot app's intelligent local routing currently supports managed Windows ML only. Local-model eligibility is separate from generic BYOK discovery. Routing decisions still occur in GitHub's cloud and local models can have different safety protections.

How to interpret the benchmark and rollout

GitHub cites a Fusion Max score of 60 on Artificial Analysis Coding Agent Index, within two points of Claude Code Opus 5.5 Max, at about half average API cost and 40% less task time. Those numbers describe the reported benchmark setup and do not establish equivalent savings for all repositories. GA applies to the core Fusion orchestration; local-model routing remains an opt-in experimental preview.

What to watch next

  • When the temporary 30% model-cost discount ends and how usage economics change.
  • Independent task-level quality and total-cost comparisons.
  • Whether local model routing graduates from preview and gains offline routing or broader app support.
  • Per-turn routing transparency and enterprise audit/attribution controls.

Still unclear

  • Benchmark economics and performance are vendor-reported and workload-dependent.
  • The limited-time discount has no explicit end date in the inspected announcement.
  • Local-model routing remains experimental and cloud-mediated.
  • Available models vary with enterprise policies and supported client/runtime.

Sources

Direct reading behind this dossier.

5 sources

Discussion

Discussion is reader-contributed. Comments are not part of the BTN dossier or its editorial evidence.

0 visible comments

Join the discussion

Keep comments useful and relevant. Reader contributions may be moderated and are not BTN editorial evidence.

Sign in to comment