HELM VHELM - Mm Star Perception - Fine-grained Perception: leaderboard
Metric: Prefix Quasi-Exact Match (%). Source: crfm.stanford.edu. 48 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro (Preview 03-25) | 74.8 |
| 2 | Llama 4 Maverick Instruct FP8 | 70.4 |
| 3 | Gemini 2.0 Flash | 68.4 |
| 4 | Gemini 2.0 Pro (Preview 02-05) | 68 |
| 5 | Gemini 2.0 Flash (Preview) | 67.6 |
| 6 | O1 (2024-12-17) | 66.4 |
| 7 | O3 (2025-04-16) | 66 |
| 8 | GPT-4.5 | 66 |
| 9 | O4 Mini (2025-04-16) | 65.2 |
| 10 | GPT-4.1 (2025-04-14) | 64.4 |
| 11 | Gemini 2.0 Flash Lite | 64 |
| 12 | Llama 4 Scout Instruct | 63.2 |
| 13 | GPT-4o (2024-05-13) | 63.2 |
| 14 | GPT-4o (2024-11-20) | 63.2 |
| 15 | Gemini 1.5 Pro (002) | 61.2 |
Interactive version: theaggregate.ai/benchmark?slug=helm-vhelm-mm-star-perception-fine-grained-perception · How It Works · Data refreshed daily, snapshot 2026-09-19.