ClawsBench - Task Success Rate (Full Scaffold): leaderboard

Task success on the 20 capability tasks of ClawsBench (2026), where an LLM agent works in stateful mock Gmail, Slack, Calendar, Docs and Drive services through the OpenClaw harness with domain skills and a meta prompt enabled, five repeats per task: the share of trials scoring at least 0.8 of the programmatic task score. Percent; higher is better.

Metric: Task success rate (%): share of trials on the 20 non-safety tasks scoring at least 0.8 of the [0, 1] task score, ClawsBench's 44 tasks over five stateful mock services (Gmail, Slack, Calendar, Docs, Drive), OpenClaw harness with domain skills and the meta prompt both enabled (full scaffolding), 5 repeats per task, state-based programmatic scoring; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 6 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Claude Opus 4.663#114
2GLM-560#276
3Gemini 3.1 Pro (Preview)58#102
4Claude Sonnet 4.656#179
5GPT-5.453#138
6Gemini 3.1 Flash Lite39#377

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

Interactive version: theaggregate.ai/benchmark?slug=clawsbench-task-success-rate-full-scaffold · How It Works · Data refreshed daily, snapshot 2026-11-22.