Mistral AI · Mistral AI
Mixtral 8x7B
SupersededReleased 2 years ago· Dec 1, 2023Language model
Landmark Apache 2.0 Mixture-of-Experts open model (8 experts, 2 active, 46B total/13B active); still popular for local deployment.
- Tier
- Open Weight
- Released
- Dec 1, 2023
Release & change timeline
1 event- Release
Mixtral 8x7B released
Apache 2.0 Mixture-of-Experts open model; outperformed GPT-3.5 and Llama 2 70B at far lower inference cost.
Part of the Mistral AI version timeline. Catch Me Up — everything that changed across AI & smart home ↗