# Diagram-to-Code on M3²Bench

Source: [Diagram-to-Code on M3²Bench](<https://www.sota2.com/research/sota/diagram-to-code-on-m3-bench>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-04-08T08:29:12.401Z

Evaluation of model performance on the Diagram-to-Code task using the M3²Bench dataset, measured by metrics including Execution Rate (Exec(%)), Structural Visualization (Svis), and Code Similarity (Scode).

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Diagram-to-Code |
| Dataset (including any stated split) | M3²Bench |
| Metric | Execution Rate |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 14 of 14 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [Execution Rate](<https://www.sota2.com/research/sota/diagram-to-code-on-m3-bench.md?metric=Execution%20Rate>): Higher is better; unit: PERCENT.

- [Structural Visualization Score](<https://www.sota2.com/research/sota/diagram-to-code-on-m3-bench.md?metric=Structural%20Visualization%20Score>): Higher is better; unit: PERCENT.

- [Code Similarity Score](<https://www.sota2.com/research/sota/diagram-to-code-on-m3-bench.md?metric=Code%20Similarity%20Score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Execution Rate | Result date | Report paper |
| --- | --- | --- | --- |
| OmniDia-7B | 94.3 | Not recorded | 69d5e0f77d0172140ec75985 |
| OmniDia-3B | 93 | Not recorded | 69d5e0f77d0172140ec75985 |
| OmniDia-7B | 92.9 | Not recorded | 69d5e0f77d0172140ec75985 |
| OmniDia-3B | 88.6 | Not recorded | 69d5e0f77d0172140ec75985 |
| Qwen2.5-VL-72B | 84.6 | Not recorded | 69d5e0f77d0172140ec75985 |
| Gemini-3-Flash | 82.4 | Not recorded | 69d5e0f77d0172140ec75985 |
| Qwen3-VL-32B | 79.8 | Not recorded | 69d5e0f77d0172140ec75985 |
| GPT-5-mini | 77.6 | Not recorded | 69d5e0f77d0172140ec75985 |
| InternVL3.5-14B | 74.8 | Not recorded | 69d5e0f77d0172140ec75985 |
| InternVL3-8B | 65.9 | Not recorded | 69d5e0f77d0172140ec75985 |
| Qwen2.5-VL-7B | 62 | Not recorded | 69d5e0f77d0172140ec75985 |
| InternVL3.5-8B | 57.4 | Not recorded | 69d5e0f77d0172140ec75985 |
| InternVL3.5-38B | 57 | Not recorded | 69d5e0f77d0172140ec75985 |
| Qwen2.5-VL-3B | 46.2 | Not recorded | 69d5e0f77d0172140ec75985 |

### OmniDia-7B

Execution Rate: 94.3. Result ID: 69d611cf47f34ce924be9eb4.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e0f77d0172140ec75985.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e0f77d0172140ec75985.md>)

Recorded evaluation configuration: {"mode":"RL","parameters":"7B"}. This record does not assert independent reproduction.

### OmniDia-3B

Execution Rate: 93. Result ID: 69d611cf47f34ce924be9eb1.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e0f77d0172140ec75985.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e0f77d0172140ec75985.md>)

Recorded evaluation configuration: {"mode":"RL","parameters":"3B"}. This record does not assert independent reproduction.

### OmniDia-7B

Execution Rate: 92.9. Result ID: 69d611cf47f34ce924be9eb3.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e0f77d0172140ec75985.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e0f77d0172140ec75985.md>)

Recorded evaluation configuration: {"mode":"SFT","parameters":"7B"}. This record does not assert independent reproduction.

### OmniDia-3B

Execution Rate: 88.6. Result ID: 69d611cf47f34ce924be9eb0.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e0f77d0172140ec75985.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e0f77d0172140ec75985.md>)

Recorded evaluation configuration: {"mode":"SFT","parameters":"3B"}. This record does not assert independent reproduction.

### Qwen2.5-VL-72B

Execution Rate: 84.6. Result ID: 69d611cf47f34ce924be9eae.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e0f77d0172140ec75985.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa773a1e209131cea71d3c.md>)

Recorded evaluation configuration: {"backbone":"Qwen2.5-VL","parameters":"72B"}. This record does not assert independent reproduction.

### Gemini-3-Flash

Execution Rate: 82.4. Result ID: 69d611cf47f34ce924be9ea8.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e0f77d0172140ec75985.md>)

Recorded evaluation configuration: {"type":"Closed-Source"}. This record does not assert independent reproduction.

### Qwen3-VL-32B

Execution Rate: 79.8. Result ID: 69d611cf47f34ce924be9eac.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e0f77d0172140ec75985.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6927db54d6b0dd3342930f66.md>)

Recorded evaluation configuration: {"backbone":"Qwen3-VL","parameters":"32B"}. This record does not assert independent reproduction.

### GPT-5-mini

Execution Rate: 77.6. Result ID: 69d611cf47f34ce924be9ea7.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e0f77d0172140ec75985.md>)

Recorded evaluation configuration: {"type":"Closed-Source"}. This record does not assert independent reproduction.

### InternVL3.5-14B

Execution Rate: 74.8. Result ID: 69d611cf47f34ce924be9eab.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e0f77d0172140ec75985.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7af01e209131cea85dc9.md>)

Recorded evaluation configuration: {"backbone":"InternVL3.5","parameters":"14B"}. This record does not assert independent reproduction.

### InternVL3-8B

Execution Rate: 65.9. Result ID: 69d611cf47f34ce924be9ea9.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e0f77d0172140ec75985.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78241e209131cea76fa5.md>)

Recorded evaluation configuration: {"backbone":"InternVL3","parameters":"8B"}. This record does not assert independent reproduction.

### Qwen2.5-VL-7B

Execution Rate: 62. Result ID: 69d611cf47f34ce924be9eb2.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e0f77d0172140ec75985.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa773a1e209131cea71d3c.md>)

Recorded evaluation configuration: {"backbone":"Qwen2.5-VL","parameters":"7B"}. This record does not assert independent reproduction.

### InternVL3.5-8B

Execution Rate: 57.4. Result ID: 69d611cf47f34ce924be9eaa.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e0f77d0172140ec75985.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7af01e209131cea85dc9.md>)

Recorded evaluation configuration: {"backbone":"InternVL3.5","parameters":"8B"}. This record does not assert independent reproduction.

### InternVL3.5-38B

Execution Rate: 57. Result ID: 69d611cf47f34ce924be9ead.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e0f77d0172140ec75985.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7af01e209131cea85dc9.md>)

Recorded evaluation configuration: {"backbone":"InternVL3.5","parameters":"38B"}. This record does not assert independent reproduction.

### Qwen2.5-VL-3B

Execution Rate: 46.2. Result ID: 69d611cf47f34ce924be9eaf.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e0f77d0172140ec75985.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa773a1e209131cea71d3c.md>)

Recorded evaluation configuration: {"backbone":"Qwen2.5-VL","parameters":"3B"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.
