HELM Robo-Reward-Bench - Dlr Sara Grid Clamp Converted Externally To Rlds: leaderboard
The "Dlr Sara Grid Clamp Converted Externally To Rlds" subtask of HELM Robo-Reward-Bench. Stanford HELM's Robo-Reward-Bench: vision-language models act as reward models over robot manipulation episodes, predicting task progress or success from video.
Metric: Absolute error. Source: crfm.stanford.edu. Saturation forecast: Around December 2026. 20 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5 Mini (2025-08-07) | 1.28 | #174 |
| 2 | GPT-4.1 Mini | 1.57 | #351 |
| 3 | Gemini 2.5 Flash | 1.61 | #242 |
| 4 | GPT-5 Nano | 1.85 | #428 |
| 5 | O1 (2024-12-17) | 2.07 | #152 |
| 6 | Gemini 2.5 Pro | 2.09 | #153 |
| 7 | GPT-4.1 (2025-04-14) | 2.23 | #250 |
| 8 | Gemini 2.0 Flash | 2.24 | #344 |
| 9 | Gemini 2.5 Flash Lite | 2.49 | #421 |
| 10 | GPT-5 | 2.57 | #104 |
| 11 | Gemini 2.0 Flash Lite | 2.67 | #453 |
| 12 | GPT-4o (2024-11-20) | 2.71 | #378 |
| 13 | Qwen2.5-Omni-7B | 3.88 | #632 |
No result here: #4 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=helm-robo-reward-bench-dlr-sara-grid-clamp-converted-externally-to-rlds · How It Works · Data refreshed daily, snapshot 2026-10-23.