Mistral: Mistral Small 4 (batch)

Mistral Small 4 (batch) is a Mistral AI model that accepts both text and image input. It offers a 262,144-token context window and allows up to 209,715 completion tokens, so long documents and lengthy outputs can fit in a single request. It supports tool calling and reasoning. Structured output support is not listed, so confirm it before relying on schema-constrained responses. Pricing is $0.075 per million input tokens and $0.30 per million output tokens. That rate makes it worth shortlisting for high-volume workloads where per-token cost drives the decision, such as mixed image and text processing or tool-driven pipelines that consume large contexts. The weak point is evidence; there is no independent benchmark coverage yet, so its quality relative to other models is unproven. Teams considering it should run their own evaluations on representative tasks against known alternatives before moving production traffic.