Mistral: Saba

Mistral: Saba

mistralai · Released Feb 17, 2025 Professional
Intelligence #10 / 576
82.0 Our Score
AA Index #285 / 378
12.1 Artificial Analysis
Input #273 / 577
$0.200 per 1M tokens
Output #270 / 577
$0.600 per 1M tokens
Context #393 / 577
32,768 tokens

Analysis Summary

Mistral Saba is a compact model from Mistral, with an intelligence index of 12.1 and moderate MMLU Pro and GPQA scores. It supports tool use and function calling, and accepts file inputs alongside text, giving it basic agentic utility. The 32K context window is restrictive for document-heavy work.

For businesses, it fits narrow, well-defined tasks such as simple content generation, classification, or lightweight automation where cost is the primary constraint. Its reasoning and coding capability are limited compared to mid-tier models, so it is not suitable for complex analysis or software engineering workflows.

At $0.20 input and $0.60 output per million tokens, it is priced attractively. Teams needing a cheap, fast model for high-volume, low-complexity tasks may find it useful, but most business workflows requiring reliability and depth will be better served by a stronger model.

Assessed June 6, 2026

Editorial notes

Mistral Saba is a lightweight, low-cost model from Mistral with tool use and function calling, suited to simple content and instruction tasks but limited in reasoning depth.

Rankings consider pricing, capabilities, benchmarks, and real-world applicability and are refreshed as new models launch. Feedback?

Performance Profile

Intelligence2.3Technical0Value7.5Content0
Intelligence 2.3/10
Technical 0/10
Content 0/10
Value 7.5/10

How Mistral: Saba compares

Mistral: Saba ranks #285 of 378 AI models we track for overall intelligence. Its 33K-token context window is larger than 32% of the models we list. At $0.20 per million input tokens it is cheaper than 53% of comparable models.

About Mistral: Saba

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..

Capabilities

Tool Use Function Calling

Performance Indices

Source: Artificial Analysis

12.1 Intelligence Index

Benchmark Scores

Intelligence

GPQA Diamond 42.4% Graduate-level scientific reasoning
HLE 4.1% Humanity's Last Exam
MMLU Pro 61.1% Multi-task language understanding
MATH 500 67.7% Mathematical problem-solving
AIME 13% Competition mathematics
SciCode 24.1% Scientific computing

Benchmark data from Artificial Analysis and Hugging Face

How does Mistral: Saba stack up?

Compare side-by-side with other professional models.

Compare Models

Model Information

OpenRouter ID mistralai/mistral-saba
Providermistralai
Model FamilyMistral
Release Date February 17, 2025
Context Length32,768 tokens
Status Active

Pricing

Token Type Cost per 1M tokens Cost per 1K tokens
Input $0.20 $0.000200
Output $0.60 $0.000600

Live Performance

Live endpoint metrics, refreshed every 30 minutes.

289ms
Best Latency (TTFT)
39.5 tok/s
Best Throughput
0/1
Active Endpoints
Available via: Mistral

Leaderboard Categories

Frequently asked questions about Mistral: Saba

How much does Mistral: Saba cost?

Mistral: Saba costs $0.20 per million input tokens and $0.60 per million output tokens.

What is the context window of Mistral: Saba?

Mistral: Saba has a context window of 32,768 tokens (33K).

What can Mistral: Saba do?

Mistral: Saba supports tool use and function calling.

Who created Mistral: Saba?

Mistral: Saba is developed by Mistral and was released on February 17, 2025.