Copilot Fusion model orchestration goes GA, Copilot app and projects launch
Octoverse says 40% of pull requests now start with an agent
Confirmed
Status
- Claim
Confirmed
- Our reporting
- High confidence
- Importance
- 3 of 5
- Last verified
- 29 October 2026
Your AI and this story
- GPT-6.2 Astra150 days after its cutoff
- Claude Opus 5.5120 days after its cutoff
- Gemini 4 Hydrogen211 days after its cutoff
- Grok 4.7150 days after its cutoff
None of these four assistants can know about it. The closest, Claude Opus 5.5, stops 120 days before it.
Key facts
- Copilot Fusion (GA): ‘evaluates what a task needs across reasoning, code generation, debugging, and tool use, then chooses how to execute across one or more models’; GitHub’s claim on the AA Coding Agent Index: Fusion Max within 2 points of Opus 5.5 Max at half the cost and 40% less time per task (vendor-reported)
- GitHub Copilot app: an agent-first surface from issue to merge that connects agents and tools across projects (specs, design discussions, telemetry analysis as well as code); Copilot projects run agent work across multiple repositories in managed parallel sessions; session continuity across surfaces (Jay Parikh’s Universe post)
- Octoverse 2026 (Oct 28): 17.9 billion pushes in the past 12 months, more than 2024 and 2025 combined; 34.3 billion commits across public and private repos; 1.3 billion merged pull requests (3.5 million a day), more than triple two years earlier
- Octoverse on agents: more than 40% of pull requests started with an agent as of September (17% in March); new repos with an agent configuration file averaged eight times the commits of new repos without one; 61% of Copilot users used more than one model in September
- Infrastructure (The New Stack, Oct 28): 7.4 billion commits in September; GitHub added about 4 million CPU cores and 160 PB of storage in 2026 to keep up
What happened
GitHub’s annual conference was built around agents: an app to direct them, a router to pick models for them, and statistics showing how much of the platform’s activity they now generate. The Octoverse numbers say activity “bent sharply upward” partway through 2026 “just as agents went from something developers were trying out to something they were working alongside”.
Why it matters
GitHub’s data is the broadest public measure of how much software work agents now start. A jump from 17% to over 40% of pull requests in six months, and commit volume high enough to force millions of new CPU cores, show agent coding moving from trial to default. Fusion also makes model choice a platform decision rather than a developer’s.
Sources
6 sources from 3 sites. Numbers match the chips in the text.
6 sources: 3 primary, 2 press, 1 reaction
Primary
- GitHub blog (Jay Parikh): GitHub Universe 2026: Evolving every layer of the stack (Oct 28)github.blog, official
- GitHub blog: Octoverse: commits, pushes reach new highs in 2026 as developers build with agents (Oct 28)github.blog, official
- GitHub blog: The bridge to agent-scale Git infrastructure (Oct 21)github.blog, official
Press
- The New Stack: GitHub debuts Copilot Fusion (Oct 28)thenewstack.io, press
- The New Stack: GitHub saw 7.4B commits in September, adding 4M CPU cores and 160 PB of storagethenewstack.io, press
Reactions
- Simon Willison: GitHub Octoverse 2026 (Oct 28)simonwillison.net, discussion
Changes
- Filed (news agent)