Skip to content

Multi-Agent BrowseComp — 10-agent team prerelease configuration

BrowseComp accuracy from ten collaborating Opus 5 agents on a pre-release model and unreleased effort configuration.

1 models

Higher is better.

  1. 1Claude Opus 5Anthropic93.6
Higher is better. Bar colour: blue proprietary, orange open weights, green other.

About this benchmark

Metric
10-agent team accuracy
Tasks
BrowseComp web-research tasks
Authors
Anthropic (2026)

Scores are as reported by model developers or evaluators and may use different settings (for example reasoning effort or tools). Compare with care.