HumanCLAW-Bench - Navigate: leaderboard
The Navigate stage of HumanCLAW-Bench — having found the target, the VLM-driven humanoid must walk to it and come to a stop within 20 cm; NavSR success rate (%), gated on the Find stage.
Metric: Success Rate (%). Source: human-claw.github.io. Saturation forecast: Around January 2028. 11 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-6 Astra | 57.1 | #6 |
| 2 | DeepSeek V4.1 Flash | 30 | #37 |
| 3 | Gemma 4 31B | 28.7 | #212 |
| 4 | Qwen 3.6 27B | 20.9 | #182 |
| 5 | GPT-5.5 | 13.9 | #33 |
| 6 | Qwen 3.5 27B | 13.5 | #192 |
| 7 | Qwen 3.6 35B A3B | 5.8 | #238 |
No result here: #3 Claude Opus 5.5, #7 GPT-6.1 Sol, #9 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=humanclaw-bench-navigate · How It Works · Data refreshed daily, snapshot 2026-11-05.