TLDRocket
Sign in

Model Efficiency

1 summarised story about Model Efficiency, each linking back to the original source. Browse all topics →

Wednesday, 7 May 2025

Medium is the new large.

Mistral AI 1 year ago

Mistral AI released Mistral Medium 3, a new language model designed to balance performance with lower costs and simpler deployment for enterprise use. The model costs $0.4 per million input tokens and $2 per million output tokens while achieving at least 90% of Claude Sonnet 3.7's performance on standard benchmarks. The model is now available on Mistral La Plateforme and Amazon SageMaker, with deployments coming to additional cloud platforms and self-hosted environments.