ClawBench (OpenClaw) - Finance: leaderboard
Metric: Success rate (%) on the 6 finance tasks of the 153-task ClawBench paper evaluation (agents in the OpenClaw browser harness on live production websites, the final state-changing request intercepted before it reaches the server, each run judged pass or fail by an Agent-as-Judge (Claude Sonnet 4.6) against a human reference trajectory); higher is better. Source: arxiv.org. Saturation forecast: Not forecast. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.6 (High) | 50 |
| 2 | Claude Haiku 4.5 | 33.3 |
| 3 | Gemini 3 Flash | 33.3 |
| 4 | Gemini 3.1 Pro (Preview) | 16.7 |
| 5 | GLM-5 | 16.7 |
| 6 | Gemini 3.1 Flash Lite | 0 |
| 7 | GPT-5.4 (High) | 0 |
Interactive version: theaggregate.ai/benchmark?slug=clawbench-openclaw-finance · How It Works · Data refreshed daily, snapshot 2026-10-07.