HM-Bench (RGB): leaderboard
HM-Bench (2026) tests multimodal LLMs on hyperspectral image understanding with 19,337 multiple-choice questions over 2,178 samples in 13 perception and reasoning tasks, given RGB-only input: a pseudo-RGB rendering of the hyperspectral cube. Accuracy (%), zero-shot; higher is better.
Metric: Average accuracy (%) over all tasks of HM-Bench's 19,337 hyperspectral multiple-choice questions (2,178 samples, 13 task categories) with RGB-only input: a pseudo-RGB rendering of the hyperspectral cube; zero-shot, one option letter as output, temperature 0, at most 64 new tokens; higher is better. Source: arxiv.org. Saturation forecast: Around 2033. 17 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GLM-4.6V-Flash | 44.46 | #1049 |
| 2 | InternVL3.5-14B | 42.88 | #1519 |
| 3 | InternVL3-14B | 42.48 | #1227 |
| 4 | InternVL3-9B | 41.74 | #1746 |
| 5 | Claude Sonnet 4.6 | 40.95 | #179 |
| 6 | Kimi-VL-A3B-Instruct | 40.24 | #2272 |
| 7 | Qwen 3 VL 4B | 40.08 | #1163 |
| 8 | GPT-5.4 Mini | 39.88 | #409 |
| 9 | GeoLLaVA-8K | 39.27 | |
| 10 | Qwen 2.5 VL 7B Instruct | 38.19 | #1919 |
| 11 | Llama-3.1-Nemotron-Nano-VL-8B-V1 | 37.54 | |
| 12 | LLaVA-Next-8B | 37.39 | #2058 |
| 13 | GeoChat | 37.03 | #3175 |
| 14 | InternVL2.5-4B | 36.74 | #2573 |
| 15 | InternVL2-8B | 36.63 | #3268 |
No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=hm-bench-rgb · How It Works · Data refreshed daily, snapshot 2026-11-22.