Mistral Small is a 22B parameter Small Language Model (SLM) designed for high efficiency and low latency workloads. It excels in tasks requiring quick responses and low computational resources, outperforming larger models while maintaining high accuracy.

Some other noteworthy features of Mistral Small include specialized RAG capabilities, strong coding performance, and multilingual support for languages including English, French, German, Spanish, and Italian.

Metric	Value
Parameter Count	22 billion
Mixture of Experts	No
Context Length	32,768 tokens
Multilingual	Yes
Quantized*	No

*Quantization is specific to the inference provider and the model may be offered with different quantization levels by other providers.