The Open-Source AI Stack
RSS
All models

Models · mistral

Mixtral 8x22B Instruct v0.1

Open Mistral AI · 2024-04-17 · Apache-2.0

Mistral's flagship open-weights sparse MoE released in April 2024 under Apache 2.0. Routes 39B active parameters per token through 2 of 8 experts, giving dense-class throughput at frontier-tier total capacity. Natively supports function calling and is multilingual across English, French, Italian, German, and Spanish.

Cost

— / Mtok input
— / Mtok output

Together AI · as of 2026-05-19

via Artificial Analysis ↗

Architecture

tokens in Embedding vocab not disclosed · mistral tokenizer × N layers Grouped-Query Attention RoPE context 65,536 tokens MoE Router 8 experts total · 2 active per token Output projection tokens out
Schema-generated from data/models.yaml. Every label is auditable against the model's sources.

Specs

Architecture
moe
Total params
141B
Active params
39B
Experts
8 total · 2 active
Context window
66K tokens
Attention
gqa
Position encoding
rope
Post-training
sft, dpo
OSI-approved
yes
Data released
no
Training code
not released

Benchmarks

Each score carries the date it was published; we never infer or interpolate missing scores.

Math

MATH 44.6 as of 2024-04-17 source ↗

Available quantizations

GGUF llama.cpp's container; the common local format, k-quants from Q2 to Q8. runs on llama.cpp, Ollama
AWQ Activation-aware 4-bit weight quantization for GPU serving. runs on vLLM, SGLang
MLX Apple MLX 4/8-bit layout for Apple silicon. runs on Apple MLX
FP8 8-bit float, frequently a native release on Hopper / Blackwell GPUs. runs on vLLM, SGLang, TensorRT-LLM

Verified via the Hugging Face model tree ↗. Community quantizations change over time; the families shown are those with published weights at audit time.

Notable innovations

  • · Sparse MoE with 2-of-8 expert routing at 22B per expert
  • · Native function calling tokens
  • · 64K context window in an Apache 2.0 release

Sources