# MedXpertQA MM — Mercor evaluation (MedXpertQA MM (Mercor))

> Medical multiple-choice reasoning over images and clinical details. This table shows Mercor-run configurations for reference and is excluded from model rankings.

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

- Category: [Multimodal & Grounded](/multimodal-grounded)
- Last updated: October 15, 2026 capture

Evaluation results published by [Mercor](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/). Published evaluation aggregates only; task contents and dataset license grants are not included.

We captured Mercor's MedXpertQA MM results on October 15, 2026. The table preserves 42 published configurations, their source identifiers, reasoning settings, and disclosed pass@1 values. These are results from Mercor's evaluation setup; we did not rerun them.

This dated source table is display only and does not enter overall or category rankings. Source configurations retain their external identity, including Pro descriptors and different effort settings.

The headline is the source page's selected Pass@1 metric. The task set, harness, tools, and grader belong to this Mercor run and may differ from another evaluator's results.

The row labels retain published error margins and sample counts when supplied. Missing margins and counts remain unreported; a sample count is not used to reconstruct a task denominator. Except where the dedicated source defines an interval, the confidence level is unreported. Equal displayed scores and overlapping margins do not establish statistical ties. Capture time does not establish individual evaluation dates.

The general methodology lists programmatic multiple-choice answer extraction and a judge of None. Its run-count label is k = 3. The linked result page takes precedence where its setup differs, as noted above.

- [Mercor MedXpertQA MM results](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/)
- [Mercor APEX benchmark index](https://www.mercor.com/apex/)
- [Mercor methodology](https://www.mercor.com/apex/methodology/)
- [Source paper](https://arxiv.org/abs/2501.18362)
- [Upstream evaluation code](https://github.com/TsinghuaC3I/MedXpertQA)
- [Upstream dataset](https://huggingface.co/datasets/TsinghuaC3I/MedXpertQA)

## About MedXpertQA MM (Mercor)

- Tasks: 200 source-reported tasks
- Format: Pass@1
- Difficulty: Source-specific evaluation
- Paper: [MedXpertQA MM evaluation results](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/)

The selected headline is Pass@1. Source model identifiers, effort settings, published margins, and reported sample counts remain attached to the source configuration. Public-set results and Mercor task extensions occupy separate tables.

MedXpertQA MM (Mercor) is currently displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.

## Leaderboard (42 configurations)

| Rank | Model | Configuration | Creator | Score |
|------|-------|---------------|---------|-------|
| 1 | [GPT-6 Astra](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · max reasoning · Published margin ±4.6 · 600 reported samples | OpenAI | 86.80% |
| 2 | [GPT-6 Astra](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · xhigh reasoning · Published margin ±4.4 · 600 reported samples | OpenAI | 86.50% |
| 3 | [Opus 5.5](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · max reasoning · Published margin ±4.4 · 600 reported samples | Anthropic | 85.70% |
| 4 | [GPT-6.1 Sol](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · max reasoning · Published margin ±4.7 · 600 reported samples | OpenAI | 85.70% |
| 5 | [Gemini 3.8 Flash](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · high reasoning · Published margin ±4.8 · 600 reported samples | Google | 84.70% |
| 6 | [Fable 5.1](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · max reasoning · Published margin ±4.7 · 600 reported samples | Anthropic | 83.70% |
| 7 | [Gemini 3.5 Flash](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · high reasoning · Published margin ±4.9 · 600 reported samples | Google | 83.50% |
| 8 | [Gemini 3.7 Flash](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · high reasoning · Published margin ±5 · 600 reported samples | Google | 83.50% |
| 9 | [Gemini 3.6 Flash](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · high reasoning · Published margin ±4.8 · 600 reported samples | Google | 82.80% |
| 10 | [GPT-6 Sol](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · max reasoning · Published margin ±5.2 · 600 reported samples | OpenAI | 81.30% |
| 11 | [Sonnet 5.5](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · max reasoning · Published margin ±5 · 598 reported samples | Anthropic | 80.70% |
| 12 | [Gemini 3.1 Pro](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · high reasoning · Published margin ±5.2 · 600 reported samples | Google | 80.20% |
| 13 | [Opus 5](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · max reasoning · Published margin ±5.2 · 600 reported samples | Anthropic | 79.80% |
| 14 | [Fable 5.1](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · high reasoning · Published margin ±5.2 · 600 reported samples | Anthropic | 79.70% |
| 15 | [GPT-5.6 Sol (Pro)](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · max reasoning · Published margin ±5.3 · 600 reported samples | OpenAI | 79.50% |
| 16 | [Qwen3.8-Max](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · xhigh reasoning · Published margin ±5.2 · 600 reported samples | Alibaba | 78.30% |
| 17 | [Muse Spark 1.1](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · xhigh reasoning · Published margin ±5.4 · 600 reported samples | Meta | 77.00% |
| 18 | [GPT-5.5](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · xhigh reasoning · Published margin ±5.4 · 600 reported samples | OpenAI | 76.70% |
| 19 | [Muse Spark 1.3](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · max reasoning · Published margin ±5.2 · 600 reported samples | Meta | 76.30% |
| 20 | [Fable 5](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · max reasoning · Published margin ±5.7 · 600 reported samples | Anthropic | 76.20% |
| 21 | [GPT-5.6 Terra](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · max reasoning · Published margin ±5.5 · 600 reported samples | OpenAI | 75.70% |
| 22 | [Muse Spark 1.2](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · xhigh reasoning · Published margin ±5.3 · 600 reported samples | Meta | 75.70% |
| 23 | [Grok 4.6](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · xhigh reasoning · Published margin ±5.7 · 600 reported samples | xAI | 75.20% |
| 24 | [GPT-5.4](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · xhigh reasoning · Published margin ±5.7 · 600 reported samples | OpenAI | 74.20% |
| 25 | [GLM-5.3-Flash](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · max reasoning · Published margin ±5.5 · 600 reported samples | Zhipu | 73.70% |
| 26 | [Kimi K3](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · max reasoning · Published margin ±5.3 · 600 reported samples | Kimi | 73.20% |
| 27 | [Opus 4.7](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · max reasoning · Published margin ±5.7 · 600 reported samples | Anthropic | 72.80% |
| 28 | [Opus 4.8](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · max reasoning · Published margin ±5.9 · 600 reported samples | Anthropic | 72.50% |
| 29 | [GPT-5.6 Luna](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · max reasoning · Published margin ±5.8 · 600 reported samples | OpenAI | 70.20% |
| 30 | [Grok 4.5](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · high reasoning · Published margin ±6.1 · 600 reported samples | xAI | 69.50% |
| 31 | [Grok 4.7](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · xhigh reasoning · Published margin ±6.1 · 578 reported samples | xAI | 69.00% |
| 32 | [Haiku 5.5](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · max reasoning · Published margin ±5.8 · 600 reported samples | Anthropic | 67.80% |
| 33 | [Gemini 3.5 Flash Lite](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · high reasoning · Published margin ±6 · 600 reported samples | Google | 66.70% |
| 34 | [Qwen3.5](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · Published margin ±6 · 600 reported samples | Alibaba | 65.80% |
| 35 | [Kimi K2.7 Code](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · high reasoning · Published margin ±5.7 · 600 reported samples | Kimi | 65.70% |
| 36 | [GPT-6 Luna](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · max reasoning · Published margin ±6.1 · 600 reported samples | OpenAI | 64.80% |
| 37 | [DeepSeek-V4.1-Flash](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · max reasoning · Published margin ±5.8 · 600 reported samples | DeepSeek | 63.50% |
| 38 | [Sonnet 5](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · max reasoning · Published margin ±6 · 600 reported samples | Anthropic | 63.00% |
| 39 | [Sonnet 4.6](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · high reasoning · Published margin ±6.2 · 600 reported samples | Anthropic | 61.70% |
| 40 | [Inkling](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · high reasoning · Published margin ±6 · 600 reported samples | Thinking Machines | 60.00% |
| 41 | [MiniMax-M3](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · high reasoning · Published margin ±6 · 600 reported samples | MiniMax | 59.70% |
| 42 | [GLM-5.3](https://www.mercor.com/apex/oss-benchmarks/oss-medxpertqa-mm-leaderboard/) | Single-shot multimodal · max reasoning · Published margin ±6.1 · 600 reported samples | Zhipu | 49.20% |

## FAQ

### What does MedXpertQA MM (Mercor) measure?

Medical multiple-choice reasoning over images and clinical details. This table shows Mercor-run configurations for reference and is excluded from model rankings.

### Which model leads the published MedXpertQA MM (Mercor) snapshot?

GPT-6 Astra currently leads the published MedXpertQA MM (Mercor) snapshot with a score of 86.80%.

### How many models are evaluated on MedXpertQA MM (Mercor)?

The October 15, 2026 capture contains 42 source configurations.

### Does MedXpertQA MM (Mercor) affect BenchLM's overall score?

Not directly. MedXpertQA MM (Mercor) is still displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.