Datasets:

ANR-MALADES
/

MediQAl

license: cc-by-4.0
task_categories:
  - question-answering
  - multiple-choice
language:
  - fr
tags:
  - medical-qa
  - open-ended-qa
  - multiple-choice-qa
pretty_name: MediQAl
size_categories:
  - 10K<n<100K
dataset_info:
  - config_name: mcqu
    features:
      - name: id
        dtype: string
      - name: clinical_case
        dtype: string
      - name: question
        dtype: string
      - name: answer_a
        dtype: string
      - name: answer_b
        dtype: string
      - name: answer_c
        dtype: string
      - name: answer_d
        dtype: string
      - name: answer_e
        dtype: string
      - name: correct_answers
        dtype: string
      - name: task
        dtype: string
      - name: medical_subject
        dtype: string
      - name: question_type
        dtype: string
    splits:
      - name: train
        num_examples: 10113
      - name: validation
        num_examples: 2561
      - name: test
        num_examples: 4343
  - config_name: mcqm
    features:
      - name: id
        dtype: string
      - name: clinical_case
        dtype: string
      - name: question
        dtype: string
      - name: answer_a
        dtype: string
      - name: answer_b
        dtype: string
      - name: answer_c
        dtype: string
      - name: answer_d
        dtype: string
      - name: answer_e
        dtype: string
      - name: correct_answers
        dtype: string
      - name: task
        dtype: string
      - name: medical_subject
        dtype: string
      - name: question_type
        dtype: string
    splits:
      - name: train
        num_examples: 5767
      - name: validation
        num_examples: 1466
      - name: test
        num_examples: 3384
  - config_name: oeq
    features:
      - name: id
        dtype: string
      - name: clinical_case
        dtype: string
      - name: cc_question_number
        dtype: string
      - name: question
        dtype: string
      - name: answer
        dtype: string
      - name: medical_subject
        dtype: string
      - name: question_type
        dtype: string
    splits:
      - name: test
        num_examples: 4969
configs:
  - config_name: mcqu
    data_files:
      - split: train
        path: mcqu/train.json
      - split: validation
        path: mcqu/validation.json
      - split: test
        path: mcqu/test.json
  - config_name: mcqm
    data_files:
      - split: train
        path: mcqm/train.json
      - split: validation
        path: mcqm/validation.json
      - split: test
        path: mcqm/test.json
  - config_name: oeq
    data_files:
      - split: test
        path: oeq/test.json

MediQAl

MediQAl is a French medical question answering dataset designed to evaluate the capabilities of language models in factual medical recall and clinical reasoning. It includes 32,603 questions sourced from French medical examinations across 41 medical subjects.

The dataset contains three tasks:

MCQU: Multiple-Choice Questions with a Unique correct answer
MCQM: Multiple-Choice Questions with Multiple correct answers
OEQ: Open-Ended Questions with Short Answers

Each question is labeled as either "Understanding" or "Reasoning", enabling analysis of the cognitive capabilities of language models.

Usage Example

from datasets import load_dataset

# Load the MCQU split
dataset = load_dataset("ANR-MALADES/MediQAl", name="mcqu")

# Explore an entry
print(dataset["train"][0])

Dataset Details

Languages: French (fr)
Domain: Medical Education
License: CC-BY-4.0
Source: French national medical examinations
Total Examples: 32,603
Medical Subjects: 41 specialties

Configurations

Config	Task Type	Answer Format
`mcqu`	Multiple choice (unique answer)	One correct option (A–E)
`mcqm`	Multiple choice (multiple answers)	Multiple correct options (A–E)
`oeq`	Open-ended question answering	Short free-text answer

Dataset Structure

Features

`mcqu` / `mcqm`

id: Question ID
clinical_case: Clinical scenario
question: The main question
answer_a to answer_e: Answer choices
correct_answers: One (mcqu) or more (mcqm) correct choices (e.g., "A,C,E")
task: Task type (mcqu / mcqm)
medical_subject: Specialty (e.g., cardiology)
question_type: "Understanding" or "Reasoning"

`oeq`

id: Question ID
clinical_case: Clinical scenario
cc_question_number: Sub-question number
question: Open-ended question
answer: Short answer
medical_subject: Specialty
question_type: "Understanding" or "Reasoning"

Dataset Characteristics

	MCQU Understanding	MCQU Reasoning	MCQU Total	MCQM Understanding	MCQM Reasoning	MCQM Total	OEQ Understanding	OEQ Reasoning	OEQ Total
Total Number of Questions	11,336	5,681	17,017	7,742	2,875	10,617	1,842	3,125	4,969
# Isolated Questions	9,126	961	10,087	6,200	343	6,543	836	179	1,015
# In-context Questions	2,210	4,720	6,930	1,542	2,532	4,074	1,006	2,946	3,954
Avg Question Length (words)	18.95	21.57	19.82	13.20	16.12	13.99	16.79	20.95	19.40
Avg Clinical Scenario Length (words)	83.50	107.67	99.97	94.87	114.77	107.24	109.71	141.28	132.19
Avg Answer Length (words)	-	-	-	-	-	-	25.26	40.24	34.68

Splits

Split	MCQU	MCQM	OEQ
Train	✅ 10,113	✅ 5,767	❌
Validation	✅ 2,561	✅ 1,466	❌
Test	✅ 4,343	✅ 3,384	✅ 4,969

Citation

@article{bazoge2026mediqal,
  title={MediQAl: A French Medical Question Answering Dataset for Knowledge and Reasoning Evaluation},
  author={Bazoge, Adrien},
  journal={Scientific Data},
  year={2026},
  publisher={Nature Publishing Group UK London}
}