Mistral Large 4 Preview Intelligence, Performance and Price Analysis (artificialanalysis.ai)

Mistral has announced the release of its latest model, Mistral Large 4 Preview, on October 6, 2026. This proprietary model is notable for its above-average performance on the Artificial Analysis Intelligence Index, scoring 38 against a median of 26, which indicates a heightened capability for reasoning and understanding complex tasks. The model supports both text and image inputs, boasts a substantial context window of 524k tokens—equivalent to about 786 A4 pages—and operates at a speed of 116 tokens per second, outperforming the median of 87 tokens per second. What makes Mistral Large 4 Preview particularly significant for the AI/ML community is its competitive pricing strategy, costing $1.36 per million input tokens and $4.18 per million output tokens, considerably lower than the median prices of $2.00 and $10.00, respectively. This affordability, combined with its extensive token generation of 200 million tokens during evaluations—a stark contrast to the median of 81 million—positions it as a compelling choice for developers seeking advanced, cost-effective AI solutions. With one trillion parameters, this model not only enhances the capabilities of generative tasks but also opens new avenues for applications in multimedia processing and complex reasoning tasks.