This is the batch-processing variant of Mistral Medium 3.1 (mistral-medium-3.1), Mistral AI's proprietary enterprise-grade model (131K-token context, multimodal text, image, and PDF input) released in August 2025. It delivers frontier-level coding, STEM reasoning, and enterprise performance at roughly 8x lower cost than top-tier models, and the batch endpoint further cuts cost via asynchronous processing for high-volume, non-latency-sensitive use.
Taxa de custo
0.2x
Contexto
131K
Lançamento
13 de ago. de 2025
Entrada
TextImageFile
Saída
Text
Suporte
RaciocínioChamada de ferramentasSaídas estruturadas