BrowseComp — Named multi-agent-swarm scaffold, distinct tools axis from single/multi-agent siblings
| # | Model | Vendor | Best score | Runs | Last seen |
|---|---|---|---|---|---|
| 1 | Kimi K2.6 | Moonshot | 86.3 | 1 | 2026-05-03 |
| 2 | Gemini 3.1 Pro | 85.9 | 1 | 2026-06-06 | |
| 3 | Claude Opus 4.6 | Anthropic | 83.7 | 1 | 2026-06-06 |
| 4 | GPT-5.4 | OpenAI | 82.7 | 1 | 2026-05-03 |
| 5 | Kimi K2.5 | Moonshot | 78.4 | 2 | 2026-06-15 |
Best tracked score per model (default configuration; source-attributed and verification-tiered). Open a model for its full benchmark surface, provenance and pricing.