Developers who care more about accuracy than cost can run several models plus a judge in one call, paying for every model involved.
Kenning's take
What it does
Fusion is a feature of OpenRouter that runs a prompt through several models at once, each with web search, then has a judge model compare the answers for agreement, contradictions and gaps before writing a final response. It is called like a single model through the openrouter/fusion slug, with preset or custom panels. OpenRouter reported that fused panels beat individual frontier models on the DRACO deep research benchmark.
- Panel of models answering in parallel
- Judge model synthesises a final answer
- Web search and fetch for each panel member
- Quality and Budget presets or custom panels
- Single API slug, OpenAI-compatible
- Try in the OpenRouter chatroom
Who it suits. Developers building research or analysis features where accuracy matters more than cost.
Key facts
- Category
- Developer tools
- Made by
- OpenRouter, Inc
- Released
- 12 June 2026
- How to get it
- OpenRouter API (model openrouter/fusion) and the OpenRouter chatroom
- Works on
- API, web chat
- API price
- sum of underlying model completions (panel plus judge)
FAQ
When did OpenRouter launch Fusion?
OpenRouter announced Fusion on 12 June 2026.
How much does Fusion cost?
Each request is billed as the sum of the panel and judge model calls.
How do I use Fusion?
Call the model slug openrouter/fusion on the OpenRouter API, or try it in the OpenRouter chatroom.
Facts from the maker's own pages and the sources listed. Checked 2026-10-11.
Similar tools

Claude Keychain
A free Mac menu bar app that keeps API keys in the macOS Keychain so Claude Code can use them without seeing them.
OpenAI Agent Builder
OpenAI's visual canvas for building multi-agent workflows, being withdrawn on 30 November 2026.
ElevenLabs Agent Workflows
Visual editor in ElevenLabs' agent platform for designing branching conversation flows for voice and chat agents.

AgentOps
Monitoring and debugging platform that records what AI agents do, what they cost and where they fail.