ClawBench (OpenClaw) - Pets: leaderboard

ClawBench (2026, NAIL Group) tests web agents on everyday write-heavy online tasks such as purchases, bookings and applications on live websites, stopping the final state-changing request and judging each run against a human reference trajectory. This board is the task success rate in percent on the pet-related tasks with agents in the OpenClaw harness; higher is better.

Metric: Success rate (%) on the 11 pets 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. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1Claude Haiku 4.518.2#587
2GLM-518.2#276
3Claude Sonnet 4.6High18.2#179
4Qwen 3.5 (ClawBench checkpoint unspecified)18.2
5Gemini 3.1 Flash Lite9.1#377
6GPT-5.4High9.1#138
7Gemini 3.1 Pro (Preview)0#102
8Gemini 3 Flash0#180

No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=clawbench-openclaw-pets · How It Works · Data refreshed daily, snapshot 2026-11-22.