SimpleVQA with tools
Multimodal · 2025-02-18
SimpleVQA's bilingual (English and Chinese) visual factuality questions answered with tool use enabled, such as code execution or image manipulation, as declared by the reporting source. The score is the correct rate over all questions under an LLM judge. Tool-free runs are scored under SimpleVQA.
Top models (higher is better)
| Model | Score |
|---|---|
| Gemini 3.7 Flash | 77.8 |
| Muse Spark 1.2 | 75.0 |
| Muse Spark Open | 73.6 |
| Muse Spark 1.1 | 73.3 |
| Kimi K3 | 71.9 |
| Claude Opus 5 | 70.3 |
| GPT-5.6 Sol | 64.7 |
| Inkling | 52.3 |