# MedXpertQA (Text) leaderboard

Knowledge benchmark

> Every figure below is reproduced as its upstream source published it: nothing is modelled, estimated, interpolated or converted. `Unavailable` means no source published the value — it is never a zero. Each value carries its evidence state and the date it was observed.

Page: https://modelchoice.dev/leaderboards/benchmarks/knowledge-med-xpert-qa-text

MedXpertQA Text. Every model the catalog carries a published MedXpertQA (Text) value for, ranked by that value.

## Benchmark record

| Field | Value |
| --- | --- |
| Category | Knowledge |
| Measure | Medical MCQ |
| Tasks | 2,450 medical multiple-choice questions |
| Difficulty | Professional medical knowledge |
| Published by | [Muse Spark Eval Methodology](https://ai.meta.com/static-resource/muse-spark-eval-methodology) |

A medical multiple-choice benchmark spanning many specialties with 10 answer options per question.

## MedXpertQA (Text) ranking

6 models with a published MedXpertQA (Text) value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.

| Rank | Model | Provider | Medical MCQ | Evidence |
| --- | --- | --- | --- | --- |
| 1 | [Gemini 3.1 Pro](https://modelchoice.dev/models/gemini-3-1-pro) | Google | 71.5 | Observed 2026-11-21 · source benchlm:models |
| 2 | [GPT-5.4](https://modelchoice.dev/models/gpt-5-4) | OpenAI | 59.6 | Observed 2026-11-21 · source benchlm:models |
| 3 | [Muse Spark](https://modelchoice.dev/models/muse-spark) | Meta | 52.6 | Observed 2026-11-21 · source benchlm:models |
| 4 | [Claude Opus 4.6](https://modelchoice.dev/models/claude-opus-4-6) | Anthropic | 52.1 | Observed 2026-11-21 · source benchlm:models |
| 5 | [Ling 3.0 Flash Sante](https://modelchoice.dev/models/ling-3-0-flash-sante) | InclusionAI | 51.0 | Observed 2026-11-21 · source benchlm:models |
| 6 | [Grok 4.20](https://modelchoice.dev/models/grok-4-20-beta) | xAI | 50.2 | Observed 2026-11-21 · source benchlm:models |

Rows are ordered by the value the source published, highest first. The source does not state whether a higher value is the better result, so this document does not either: for a benchmark that measures a rate of failure, read the table from the bottom.
