Mixtral 8x7B

Mixture-of-experts model with efficient inference

Overview

Mixture-of-experts model with efficient inference The model has 46.7B (12.9B active) parameters. It is an open-weight model released under the Apache 2.0 licence.

Key Capabilities

Code Generation — suited for tasks requiring code generation proficiency
Reasoning — suited for tasks requiring reasoning proficiency
Multilingual — suited for tasks requiring multilingual proficiency

Recommended Use Cases

Multilingual Tasks
Efficient Large Model
Self Hosting

Technical Specifications

Context Window: 33K tokens
Parameters: 46.7B (12.9B active)
Open Weight: Yes
Licence: Apache 2.0
Release Date: December 11, 2023

Provider

Mixtral 8x7B is developed by Mistral AI. For the latest information, visit the official documentation.