MeasureBench Real-world: leaderboard
Each model reads the value and unit off 1,272 photographs of measuring instruments (dials, digital displays, linear scales, composites). The score is the percentage of answers whose value falls in the annotated interval and whose unit matches; unlike chart or OCR benchmarks the hard part is locating a pointer or meniscus between tick marks.
Metric: Accuracy (%). Source: raw.githubusercontent.com. Saturation forecast: Around 2029. 13 models tracked. Updated 17 Nov 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Qwen 3 VL 235B A22B Instruct | 22.6 | #540 | |
| 2 | GPT-5 Mini | 22 | #359 | |
| 3 | GPT-5 | 19.8 | #177 | |
| 4 | Qwen 2.5 VL 7B Instruct | 14.6 | #1893 | |
| 5 | Qwen 2.5 VL 72B Instruct | 14.5 | #849 | |
| 6 | Claude Opus 4.1 | Non-reasoning | 14.3 | #279 |
| 7 | InternVL3.5-38B | Non-reasoning | 12.9 | #1017 |
| 8 | Claude Sonnet 4 | 12.6 | #411 | |
| 9 | Qwen 2.5 VL 32B Instruct | 11.7 | #1074 | |
| 10 | Mistral Medium 3.1 | 10.6 | #865 | |
| 11 | InternVL3.5-8B | Non-reasoning | 9.7 | #1588 |
| 12 | Mistral Small 3.2 | 8.5 | #1645 | |
| 13 | Grok 4 | 7.5 | #315 |
No result here: #1 Claude Fable 5.2, #2 GPT-6.2 Astra, #3 Gemini 4 Carbon.
Interactive version: theaggregate.ai/benchmark?slug=measurebench-real-world · How It Works · Data refreshed daily, snapshot 2026-11-17.