DynaMath
We show this table for reference; we do not rank on it.
A multimodal benchmark for dynamic mathematical reasoning over visual and structured inputs.
Benchmark score on DynaMath — November 13, 2026
We compile the DynaMath rows from provider self-reports and secondary reports. GPT-5.2 leads the table at 86.8%, followed by Qwen3.5 397B (86.3%) and Qwen3.6-27B (85.6%). We do not use these results to rank models overall.
GPT-5.2
OpenAI
Another lab’s comparison tableQwen3.5 397B
Alibaba
Reported by AlibabaQwen3.6-27B
Alibaba
Reported by Alibaba7 modelsMultimodal & GroundedCurrentDisplay onlyUpdated November 13, 2026
Benchmark score table (7 models)
None of the 7 results on this page has a documented run by the benchmark or an independent evaluator: 3 reported by the labs that built the models and 4 reported in another publisher’s document.
| Rank | Model / configuration | Score | Who ran it | Parameters (B) | Open / closed |
|---|---|---|---|---|---|
| 1 | GPT-5.2OpenAIAnother lab’s comparison table | 86.8% | Another lab’s comparison table | Not reported | Closed |
| 2 | Qwen3.5 397BAlibabaReported by Alibaba | 86.3% | Reported by Alibaba | Not reported | Open |
| 3 | Qwen3.6-27BAlibabaReported by Alibaba | 85.6% | Reported by Alibaba | Not reported | Open |
| 4 | Gemini 3 ProGoogleAnother lab’s comparison table | 85.1% | Another lab’s comparison table | Not reported | Closed |
| 5 | Kimi K2.5Moonshot AIAnother lab’s comparison table | 84.4% | Another lab’s comparison table | Not reported | Open |
| 6 | Claude Opus 4.5AnthropicAnother lab’s comparison table | 79.7% | Another lab’s comparison table | Not reported | Closed |
| 7 | LFM3-VL-2BLiquidAIReported by LiquidAI | 79.2% | Reported by LiquidAI | Not reported | Open |
- Lab-reported
- Benchmark-run
- Independent
Among the reported DynaMath rows, GPT-5.2 is first at 86.8%. The third row is 1.2 points behind. The broader top-10 range is 7.6 points, so many of the published results sit in a relatively narrow band.
7 models have been evaluated on DynaMath. The benchmark falls in the Multimodal & Grounded category. DynaMath is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.
About DynaMath
Year
2026
Tasks
Dynamic visual math problems
Format
Multimodal mathematical reasoning
Difficulty
Advanced multimodal mathematics
DynaMath is intended to probe whether models can track changing mathematical structure in multimodal settings rather than solving static text-only equations.
Freshness and provenance
Version
DynaMath 2026
Refresh cadence
Quarterly
Staleness state
Current
Question availability
Public benchmark set
BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.
Questions
What does DynaMath measure?
A multimodal benchmark for dynamic mathematical reasoning over visual and structured inputs.
Which model scores highest on DynaMath?
GPT-5.2 by OpenAI currently leads with a score of 86.8% on DynaMath.
How many models are evaluated on DynaMath?
7 AI models have published results on DynaMath in the BenchLM catalog.
Compare top models on DynaMath
Know when it’s worth switching models
The model to choose, the cheaper alternative, and the release we would wait on.
Read a sample issueJoin 7,000+ subscribers.
One email each week. Unsubscribe anytime.