Counts toward the leader of leaders
Best for agents and tools
Calling functions and tools, using MCP servers, searching the web. 3 fresh boards.
The wall
Ranks on each board in this category.
| Agents and tools | |||
|---|---|---|---|
| Model · podium score | BFCL | MCP Atlas | Search Arena |
1 GPT-5.6 Sol33.3 | |||
1 Muse Spark 1.133.3 | |||
3 Claude Opus 4.533.3 | |||
| 4Claude Fable 533.3 | |||
| 5Claude Fable 5.130.0 | |||
| 6Claude Opus 4.630.0 | |||
| 7Claude Sonnet 4.530.0 | |||
| 8Claude Opus 526.7 | |||
| 9GPT-5.526.7 | |||
10 Gemini 3 Pro26.7 | |||
#1 with the vendor’s logo23 podium7 top 10 stale board, not countedArrow keys move, Enter opens
Rank over time
Category rank at each month’s end.
Boards
Berkeley Function Calling Leaderboard
#1 Claude Opus 4.5, 77.5%
MCP Atlas
#1 Muse Spark 1.1, 88.1%
LMArena Search Arena
#1 GPT-5.6 Sol, 1,257
Scores as published by each board on the capture date. Model names and logos belong to their owners; logos via logo.dev.