Input: $0.6, Output: $0.6

Cost per million tokens

For enterprise pricing and custom weights or models

Mixtral 8x7b

Mixtral 8x7b is a large language model (LLM) created by Mistral AI. It's known for being efficient and powerful. Here's a quick rundown of its key features:

  1. Efficient: Mixtral 8x7b is a sparse model, meaning it only uses a portion of its parameters at a time. This makes it faster and cheaper to run than some other LLMs.

  2. Powerful: Despite its efficiency, Mixtral 8x7b performs well on many benchmarks, even exceeding some larger models.

  3. Multilingual: It can understand and respond in English, French, Italian, German, and Spanish.

  4. Open-source: The code is available for anyone to use and modify.