UniClawBench - Average Score: leaderboard
Metric: Average checkpoint score (%; mean completed share of the task checkpoints, English and Chinese subsets averaged) on 400 bilingual real-world tasks run live in Docker containers under the OpenClaw framework (v2026.3.11), with up to two simulated user follow-ups and a hidden GPT-5.4 Codex supervisor that grades step checkpoints. Source: arxiv.org. Saturation forecast: Around December 2026. 10 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5.4 | 77.4 |
| 2 | Claude Sonnet 4.6 | 76.3 |
| 3 | Qwen 3.5 Plus | 73.1 |
| 4 | Gemini 3.1 Pro (Preview) | 72.7 |
| 5 | Kimi K2.6 | 70.9 |
| 6 | Claude Opus 4.8 | 70.2 |
| 7 | Gemini 3 Flash | 69.8 |
| 8 | GPT-5.4 Mini | 61.6 |
| 9 | Gemini 3.1 Flash Lite | 54.3 |
| 10 | GPT-4.1 | 48.8 |
Interactive version: theaggregate.ai/benchmark?slug=uniclawbench-average-score · How It Works · Data refreshed daily, snapshot 2026-09-29.