GitHub officially releases Copilot Fusion, automatically combining multiple AI models based on the task
At GitHub Universe 2026, GitHub announced the official release of GitHub Copilot Fusion, a feature that selects a single AI model or a combination of multiple models to execute based on task characteristics. The original article presents three collaboration modes, one of which has a low-cost model draft a plan, a quality-check model verify it and decide whether to escalate to a stronger model, while another has models from different model families independently review the draft and lets the drafting model make one revision. The feature offers three load levels—Fusion Standard, Fusion Hight, and Fusion Max—to balance quality and cost, plus a Fusion Fast that users cannot specify; Fusion Standard also supports locally installed MAI Code 1.1 Flash or Qwen3.8 Flash Next models to execute tasks, in public preview.
Why read this
AI score 62
The original article breaks down Copilot Fusion's three model collaboration modes and four load levels, which can be used to assess how multi-model orchestration weighs cost against quality.
Topics
Body · AI translation
Google Ad Server Settings
Simple Analytics Placeholder event function
×
End of Header
End of superbanner /3002452/Pbk_Rect1_300x300
/3002452/Pbk_Rect2_300x250
GitHub officially releases "Copilot Fusion." High performance and low cost through optimal AI model selection according to the task. GitHub Universe 2026, November 2, 2026
At "GitHub Universe 2026," its annual event held over two days, October 28 and 29, in San Francisco, USA, GitHub announced the official release of its new feature "GitHub Copilot Fusion."
Copilot Fusion is a multi-model feature that executes by combining a single or multiple AI models according to the characteristics of a given task.
Copilot Fusion selects one of the following three patterns according to a given task.
The AI model selected as the efficient model creates a draft solution, and the AI model for quality verification verifies it and decides whether to escalate to a higher-capability AI model.
Critique
First, one AI model creates a draft solution, and an AI model from a different model family independently reviews it. When the review returns points of concern, based on those points, the AI model that created the draft revises the solution only once.
In selecting the above patterns, load levels are provided for adjusting how to balance quality and cost.
Fusion Standard
Balances cost, quality, and latency. For everyday work.
Fusion Hight
Balanced orchestration with moderate token savings. For large-scale coding tasks with clear scope, such as fixing complex bugs or making changes spanning multiple files.
Fusion Max
Achieves state-of-the-art performance while considering token efficiency. For tasks requiring the deepest possible reasoning and the most thorough analysis.
Fusion Fast
Selected in scenarios requiring a high-speed model. Users cannot specify it.
Additionally, as a public preview, in Fusion Standard it is also possible to select a locally installed MAI Code 1.1 Flash model or Qwen3.8 Flash Next model to execute tasks.
Read also
Machine Learning / AI Development Tools GitHub Microsoft
≪ Previous article: Oracle announces "Oracle App Composer," which lets employees automatically create business apps with AI and share them on Oracle Cloud. Oracle AI World 2026
After working as an editor at an IT magazine and a publisher of an online media outlet, became independent. Started Publickey in 2009.
(Detailed profile)
Would you like to check the latest news from Publickey?
GitHub officially releases "Copilot Fusion." High performance and low cost through optimal AI model selection according to the task. GitHub Universe 2026
Oracle announces "Oracle App Composer," which lets employees automatically create business apps with AI and share them on Oracle Cloud. Oracle AI World 2026
Oracle Cloud announces adoption of Google TPU, to begin offering in 2027. Oracle AI World 2026
JavaScript runtime "Bun" becomes a Prettier-compatible formatter more than 300 times faster. Prettier configuration files can also be used as-is
AWS and Google Cloud open-source versions of "PyTorch" for AWS Trainium and Google Cloud TPU. Taking on the challenge of migrating from NVIDIA
"DuckDB 2.0 LTS" officially released. Client/server functionality via the Quack protocol, speedups through asynchronous processing, and more
Cloudflare announces "tsrs," a port of TypeScript to Rust. Up to 70% less memory and 8 times faster than TypeScript 7. Plans to support all TypeScript features including LSP in the future
To provide "Cloudflare OS" as a managed service. Connections with GitHub, enhanced integration with Gmail, and more
AWS announces "AWS Lambda Web Functions," which can deploy and run Node.js frameworks and apps such as Next.js and Hono on AWS Lambda as-is
Our efforts to ensure readers can read articles with peace of mind
Simple Analytics 100% privacy-first analytics
- script for simple analytics events --
Global site tag (gtag.js) - Google Analytics