About

Mistral Saba is a 24B-parameter language model specifically designed for the Middle East and South Asia, delivering accurate and contextually relevant responses while maintaining efficient performance. Trained on curated regional datasets, it supports multiple Indian-origin languages—including Tamil and Malayalam—alongside Arabic. This makes it a versatile option for a range of regional and multilingual applications. Read more at the blog post [here](https://mistral.ai/en/news/mistral-saba)

Model Family

Mistral Small 4 (Reasoning)2026-03-16 Mistral Small 4 (Non-reasoning)2026-03-16 Mistral: Mistral Small 42026-03-16 Mistral: Mistral Small Creative2025-12-16 Mistral Large 32025-12-02 Mistral: Mistral Large 3 25122025-12-01 Mistral Medium 3.12025-08-12 Mistral Small 3.22025-06-20

Benchmarks

MMLU-Pro

61.1%

GPQA Diamond

42.4%

HLE

4.1%

LiveCodeBenchNot evaluated

SciCode

24.1%

TerminalBench HardNot evaluated

MATH-500

67.7%

AIME

13.0%

AIME 2025Not evaluated

IFBenchNot evaluated

Long Context RecallNot evaluated

Tau2Not evaluated

Market AverageTop Score

Mistral Saba

About

Model Family

Pricing

Cost Calculator

vs. Similar Models

Performance

Benchmarks

Open Source

Quick Compare

Similar Models