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(opens in new tab)
Modalities
In / Out Price
$0.20 / $0.60per 1M
Context
33K
Released
Feb 17, 2025
Knowledge Cutoff
Sep 2024
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.
Saba costs $0.20/M input tokens and $0.60/M output tokens, with separate rates for Cache Read at $0.02/M tokens.
Saba has a 32,768 token context window.
Yes. Saba accepts tools and tool_choice for function calling. It also supports structured outputs via a JSON schema in response_format.
Saba accepts text and files such as PDFs as input and returns text.
Mistral Medium 3.5, Mistral Small 4, Ministral 3 14B 2512 and 14 more are other text models from Mistral AI.
Saba was released on February 17, 2025. Its knowledge cutoff is September 30, 2024.