# DynaMath

> A multimodal benchmark for dynamic mathematical reasoning over visual and structured inputs.

Canonical page: https://benchlm.ai/benchmarks/dynamath

- Category: [Multimodal & Grounded](/multimodal-grounded)
- Last updated: September 27, 2026

## About DynaMath

- Year: 2026
- Tasks: Dynamic visual math problems
- Format: Multimodal mathematical reasoning
- Difficulty: Advanced multimodal mathematics
- Paper: [Qwen3.6 launch benchmarks](https://qwen.ai/blog?id=qwen3.6)

DynaMath is intended to probe whether models can track changing mathematical structure in multimodal settings rather than solving static text-only equations.

DynaMath is currently displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.

## Leaderboard (1 models)

| Rank | Model | Creator | Score |
|------|-------|---------|-------|
| 1 | [Qwen3.6-27B](/models/qwen3-6-27b) | Alibaba | 85.6% |

## FAQ

### What does DynaMath measure?

A multimodal benchmark for dynamic mathematical reasoning over visual and structured inputs.

### Which model scores highest on DynaMath?

Qwen3.6-27B by Alibaba currently leads with a score of 85.6% on DynaMath.

### How many models are evaluated on DynaMath?

1 AI models have been evaluated on DynaMath on BenchLM.

### Does DynaMath affect BenchLM's overall score?

Not directly. DynaMath is still displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.