Mixtral 8x22B
LegacyPremiumMistral's large MoE model with 8 expert groups and 22B each.
Strong at code, math, and multilingual tasks with efficient sparse inference. Superseded by Mistral Large series.
▼
Mixtral 8x22B is very affordable at $2.00/M tokens input and $6.00/M tokens output. Cached input reads are priced at $0.20/M tokens, making repeated context cheaper. The model supports a 66K context window.
Benchmarks
Specifications
Performance
Providers
Misc
Compare with
Frequently Asked Questions about Mixtral 8x22B▼
When was Mixtral 8x22B released?
▼
Mixtral 8x22B was released on April 17, 2024.
Who created Mixtral 8x22B?
▼
Mixtral 8x22B was created by Mistral.
How much does Mixtral 8x22B cost?
▼
Mixtral 8x22B costs $2.00/M input tokens and $6.00/M output tokens.
What is Mixtral 8x22B API pricing?
▼
The Mixtral 8x22B API is priced at $2.00 per million input tokens and $6.00 per million output tokens. You can access Mixtral 8x22B through idapt.app alongside 200+ other AI models in one workspace.
Is Mixtral 8x22B a reasoning model?
▼
No, Mixtral 8x22B is not a reasoning model. It responds directly without an extended internal thinking step. For tasks requiring deep chain-of-thought reasoning, consider models like Claude, o-series from OpenAI, or Gemini Thinking.
Does Mixtral 8x22B support image or vision input?
▼
No, Mixtral 8x22B does not currently support image or vision input. It is a text-only model.
Does Mixtral 8x22B support audio?
▼
No, Mixtral 8x22B does not support audio input.
What is the context window of Mixtral 8x22B?
▼
Mixtral 8x22B has a context window of 66K tokens, meaning it can process approximately 49,152 words of context at once.
Is Mixtral 8x22B open source?
▼
Mixtral 8x22B is a proprietary model developed by Mistral. The model weights and training data are not publicly available.
Where can I use Mixtral 8x22B?
▼
You can chat with Mixtral 8x22B on idapt.app — no separate API key required. idapt is an AI workspace with 200+ models, agents, computers, and tasks. You can also access Mixtral 8x22B through Mistral's own API or platform.