Mistral Batch API

Mistral AI · 1 year ago

Mistral introduced a batch API that processes high-volume requests asynchronously at 50% lower cost than synchronous API calls. The batch API is available today on La Plateforme for all Mistral models, with a usage limit of 1 million ongoing requests per workspace. Developers building applications prioritizing data volume over real-time responses can now reduce inference costs for tasks like document summarization, sentiment analysis, and vector embeddings.