Mistral: Mistral Small 3.1 24B

Mistral Small 3.1 24B is a text-and-image input model from Mistral AI with a 128,000-token context window and a matching 128,000-token completion ceiling. It does not support tool use, reasoning modes, or structured output, so workflows that depend on those features will need a different option. Benchmark coverage here is thin, with a blended score of 26.8 drawn from a single benchmark, so treating that figure as definitive would be premature. On price, at $0.351 per million input tokens and $0.555 per million output tokens, it sits in the budget-to-mid range. Users who need multimodal input and long-context handling at a modest cost, and who can work without tool calling or structured output, are the most natural fit. Those running high-volume pipelines may find the trade-offs acceptable, but buyers prioritizing well-validated performance should wait for broader benchmark coverage before committing.