🤖 Agent Leaderboard
Tool use, multi-step planning, and autonomous task completion.
2 models
| # | Model | Composite | Speed |
|---|---|---|---|
| 1 | Claude Opus 4 Anthropic | 97.8 | 45 t/s |
| 2 | Command R+ Cohere | 60.8 | 52 t/s |
Tool use, multi-step planning, and autonomous task completion.
| # | Model | Composite | Speed |
|---|---|---|---|
| 1 | Claude Opus 4 Anthropic | 97.8 | 45 t/s |
| 2 | Command R+ Cohere | 60.8 | 52 t/s |