AA-Robotics v0.1: leaderboard
Artificial Analysis AA-Robotics v0.1 research preview: frontier models inside coding-agent harnesses (Codex, Claude Code, Antigravity) at High reasoning effort control a physical 6-axis Hiwonder NexArm robot arm via HTTP APIs and live camera feeds across 8 tabletop manipulation tasks. Unweighted mean task score (0-100); higher is better.
Metric: Mean Task Score (0-100). Source: artificialanalysis.ai. Saturation forecast: Around May 2027. 5 models tracked. Updated 21 Nov 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | GPT-6 Astra | High | 69.38 | #10 |
| 2 | GPT-6.1 Sol | High | 64.62 | #15 |
| 3 | Gemini 4 Argon | High | 55.25 | #19 |
| 4 | Claude Opus 5.5 | High | 48.75 | #9 |
| 5 | Claude Fable 5.1 | High | 46.25 | #20 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=aa-robotics-v0-1 · How It Works · Data refreshed daily, snapshot 2026-11-21.