ThermEval - Temperature Estimation at 6 ft: leaderboard

Metric: Mean absolute error (degrees Celsius, 0 or more) on Task 7, body-region temperature from ThermEval-D images taken at 6 ft (426 questions); zero-shot with fixed one-word or one-number prompts, a single forward pass; Gemini 2.5 models parse free-form answers into the label or number only when the output deviates from the format; lower is better. Source: arxiv.org. Saturation forecast: Around May 2027. 23 models tracked.

Top models

#ModelScoreOverall rank
1Gemini 3 Pro0.74#77
2Qwen 2.5 VL 32B Instruct0.89#443
3Qwen 2.5 VL 7B Instruct0.93#643
4InternVL3-38B0.97#395
5Gemini 3 Flash1#93
6Qwen 2 VL 7B Instruct1#816
7InternVL3-8B1.01#606
8InternVL3-14B1.03#494
9Llama 3.2 11B Instruct1.74#1112
10Gemini 2.5 Pro1.95#145
11Gemini 2.5 Flash2.36#237
12MiniCPM-V-2.64.45#825

No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.

Interactive version: theaggregate.ai/benchmark?slug=thermeval-temperature-estimation-at-6-ft · How It Works · Data refreshed daily, snapshot 2026-10-11.