Mistral: Mistral Medium 3.1

Mistral Medium 3.1 from MistralAI is capable of processing text, images, and files within a context length of up to 131,072 tokens; however, it does not support reasoning or structured output. While the model can utilize external tools, its benchmark score in the AI Index stands at 24.2 out of 100 across three independent benchmarks, with specific scores of 33.8 for coding and 10.2 for agentic tasks. The pricing is $0.4 per million input tokens and $2.0 per million output tokens. This model should be considered by users who need a versatile tool that can handle multiple modalities but do not require advanced reasoning or structured outputs. Given its blended benchmark score of 21.7, it performs adequately across the tested domains, making it a cost-effective choice for those looking to integrate text and file processing into their workflows.