AgentBench FC: leaderboard
Function-calling agent benchmark across ALFWorld, database, knowledge-graph, operating-system, and WebShop environments, requiring tool calls that complete multi-step tasks.
Metric: Avg Success Rate (%). Source: github.com. Saturation forecast: Around January 2027. 25 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Sonnet 4.5 | 58.9 | #142 |
| 2 | Claude Sonnet 4.5 (Thinking) | 58.3 | #142 (Claude Sonnet 4.5) |
| 3 | Claude Sonnet 4 (Thinking) | 58.2 | #199 (Claude Sonnet 4) |
| 4 | Claude Sonnet 4 | 57.4 | #199 |
| 5 | Claude 3.7 Sonnet | 53.2 | #249 |
| 6 | GPT-5 | 52.2 | #98 |
| 7 | Claude 3.7 Sonnet (Thinking) | 50 | #249 (Claude 3.7 Sonnet) |
| 8 | DeepSeek R1 | 49.3 | #251 |
| 9 | O3 Mini | 40.9 | #272 |
| 10 | Qwen 2.5 72B Instruct | 40.8 | #454 |
| 11 | O4 Mini | 39.7 | #180 |
| 12 | GPT-4o | 39.6 | #343 |
| 13 | Qwen 2.5 32B Instruct | 37.2 | #498 |
| 14 | DeepSeek V3 | 36.1 | #321 |
| 15 | Qwen 2.5 14B Instruct | 27.2 | #647 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=agentbench-fc · How It Works · Data refreshed daily, snapshot 2026-10-20.