# Health Dialogue on HealthBench

Source: [Health Dialogue on HealthBench](<https://www.sota2.com/research/sota/health-dialogue-on-healthbench>)

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

Data updated: 2026-06-02T10:46:50.407Z

Evaluation of model performance on conversational health dialogue tasks using the HealthBench dataset, measured by accuracy.

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 | Health Dialogue |
| Dataset (including any stated split) | HealthBench |
| Metric | Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 17 of 17 |
| 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

- [Accuracy](<https://www.sota2.com/research/sota/health-dialogue-on-healthbench.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Official Instruct Model | 44.92 | Not recorded | 6a1e640adbd7150506f8564a |
| GPT-5.4 (High) | 29.39 | Not recorded | 6a1e640adbd7150506f8564a |
| GLM-4.7 & ANDES | 25.87 | Not recorded | 6a1e640adbd7150506f8564a |
| GLM-4.7 (Scaffold-only) | 16.74 | Not recorded | 6a1e640adbd7150506f8564a |
| Opus-4.7 (xHigh) | 16.23 | Not recorded | 6a1e640adbd7150506f8564a |
| Gemini-3.1-Pro | 10.93 | Not recorded | 6a1e640adbd7150506f8564a |
| Opus-4.6 (1M) | 9.73 | Not recorded | 6a1e640adbd7150506f8564a |
| GPT-5.2 | 9.33 | Not recorded | 6a1e640adbd7150506f8564a |
| Opus-4.6 | 8.58 | Not recorded | 6a1e640adbd7150506f8564a |
| MiniMax-M2.5 | 7.66 | Not recorded | 6a1e640adbd7150506f8564a |
| Base Model (Qwen3-1.7B) | 7.54 | Not recorded | 6a1e640adbd7150506f8564a |
| MiniMax-M2.1 | 7.54 | Not recorded | 6a1e640adbd7150506f8564a |
| Qwen3-Max | 7.54 | Not recorded | 6a1e640adbd7150506f8564a |
| Kimi-K2-Thinking | 7.54 | Not recorded | 6a1e640adbd7150506f8564a |
| GPT-5.1-Codex-Max | 7.54 | Not recorded | 6a1e640adbd7150506f8564a |
| GLM-4.7 (OpenCode) | 7.54 | Not recorded | 6a1e640adbd7150506f8564a |
| Sonnet-4.5 | 0 | Not recorded | 6a1e640adbd7150506f8564a |

### Official Instruct Model

Accuracy: 44.92. Result ID: 6a1eb47cfbc373592c19a21a.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e640adbd7150506f8564a.md>)

Recorded evaluation configuration: {"Evaluation Category":"Official Instruct Model"}. This record does not assert independent reproduction.

### GPT-5.4 (High)

Accuracy: 29.39. Result ID: 6a1eb47cfbc373592c19a225.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e640adbd7150506f8564a.md>)

Recorded evaluation configuration: {"Evaluation Category":"Agent-Post-Trained","Base Model":"Qwen3-1.7B"}. This record does not assert independent reproduction.

### GLM-4.7 & ANDES

Accuracy: 25.87. Result ID: 6a1eb47cfbc373592c19a22a.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e640adbd7150506f8564a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a1e640adbd7150506f8564a.md>)

Recorded evaluation configuration: {"Evaluation Category":"Proposed Method Comparison","Base Model":"Qwen3-1.7B"}. This record does not assert independent reproduction.

### GLM-4.7 (Scaffold-only)

Accuracy: 16.74. Result ID: 6a1eb47cfbc373592c19a229.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e640adbd7150506f8564a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a1e640adbd7150506f8564a.md>)

Recorded evaluation configuration: {"Evaluation Category":"Proposed Method Comparison","Base Model":"Qwen3-1.7B"}. This record does not assert independent reproduction.

### Opus-4.7 (xHigh)

Accuracy: 16.23. Result ID: 6a1eb47cfbc373592c19a227.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e640adbd7150506f8564a.md>)

Recorded evaluation configuration: {"Evaluation Category":"Agent-Post-Trained","Base Model":"Qwen3-1.7B"}. This record does not assert independent reproduction.

### Gemini-3.1-Pro

Accuracy: 10.93. Result ID: 6a1eb47cfbc373592c19a224.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e640adbd7150506f8564a.md>)

Recorded evaluation configuration: {"Evaluation Category":"Agent-Post-Trained","Base Model":"Qwen3-1.7B"}. This record does not assert independent reproduction.

### Opus-4.6 (1M)

Accuracy: 9.73. Result ID: 6a1eb47cfbc373592c19a226.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e640adbd7150506f8564a.md>)

Recorded evaluation configuration: {"Evaluation Category":"Agent-Post-Trained","Base Model":"Qwen3-1.7B"}. This record does not assert independent reproduction.

### GPT-5.2

Accuracy: 9.33. Result ID: 6a1eb47cfbc373592c19a223.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e640adbd7150506f8564a.md>)

Recorded evaluation configuration: {"Evaluation Category":"Agent-Post-Trained","Base Model":"Qwen3-1.7B"}. This record does not assert independent reproduction.

### Opus-4.6

Accuracy: 8.58. Result ID: 6a1eb47cfbc373592c19a222.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e640adbd7150506f8564a.md>)

Recorded evaluation configuration: {"Evaluation Category":"Agent-Post-Trained","Base Model":"Qwen3-1.7B"}. This record does not assert independent reproduction.

### MiniMax-M2.5

Accuracy: 7.66. Result ID: 6a1eb47cfbc373592c19a21d.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e640adbd7150506f8564a.md>)

Recorded evaluation configuration: {"Evaluation Category":"Agent-Post-Trained","Base Model":"Qwen3-1.7B"}. This record does not assert independent reproduction.

### Base Model (Qwen3-1.7B)

Accuracy: 7.54. Result ID: 6a1eb47cfbc373592c19a21b.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e640adbd7150506f8564a.md>)

Recorded evaluation configuration: {"Evaluation Category":"Zero-Shot"}. This record does not assert independent reproduction.

### MiniMax-M2.1

Accuracy: 7.54. Result ID: 6a1eb47cfbc373592c19a21e.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e640adbd7150506f8564a.md>)

Recorded evaluation configuration: {"Evaluation Category":"Agent-Post-Trained","Base Model":"Qwen3-1.7B"}. This record does not assert independent reproduction.

### Qwen3-Max

Accuracy: 7.54. Result ID: 6a1eb47cfbc373592c19a21f.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e640adbd7150506f8564a.md>)

Recorded evaluation configuration: {"Evaluation Category":"Agent-Post-Trained","Base Model":"Qwen3-1.7B"}. This record does not assert independent reproduction.

### Kimi-K2-Thinking

Accuracy: 7.54. Result ID: 6a1eb47cfbc373592c19a220.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e640adbd7150506f8564a.md>)

Recorded evaluation configuration: {"Evaluation Category":"Agent-Post-Trained","Base Model":"Qwen3-1.7B"}. This record does not assert independent reproduction.

### GPT-5.1-Codex-Max

Accuracy: 7.54. Result ID: 6a1eb47cfbc373592c19a221.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e640adbd7150506f8564a.md>)

Recorded evaluation configuration: {"Evaluation Category":"Agent-Post-Trained","Base Model":"Qwen3-1.7B"}. This record does not assert independent reproduction.

### GLM-4.7 (OpenCode)

Accuracy: 7.54. Result ID: 6a1eb47cfbc373592c19a228.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e640adbd7150506f8564a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a1e640adbd7150506f8564a.md>)

Recorded evaluation configuration: {"Evaluation Category":"Proposed Method Comparison","Base Model":"Qwen3-1.7B"}. This record does not assert independent reproduction.

### Sonnet-4.5

Accuracy: 0. Result ID: 6a1eb47cfbc373592c19a21c.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e640adbd7150506f8564a.md>)

Recorded evaluation configuration: {"Evaluation Category":"Agent-Post-Trained","Base Model":"Qwen3-1.7B"}. This record does not assert independent reproduction.

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