Mixtral Models — Hardware Requirements
7 Mixtral models from Mistral AI and the community, from the smallest that runs in 0.5 GB of VRAM at Q4 up to 140.6B parameters. Every row links to full quantization tables and GPU compatibility.
All Mixtral Models by Size
| Model | Params | Runs from (Q4) | Context | Publisher | Quant downloads |
|---|---|---|---|---|---|
| Tiny Mixtral | 247M | 0.5 GB | 131K | ||
| Mixtral 8x7B Instruct v0.1 | 46.7B | 30.5 GB | 33K | ||
| Mixtral 8x7B v0.1 | 46.7B | 30.5 GB | 33K | ||
| Mixtral 34Bx2 MoE 60B | 60.8B | 40.8 GB | 200K | ||
| Mixtral 8x22B v0.1 | 140.6B | 88.4 GB | 66K | ||
| Mixtral 8x22B Instruct v0.1 | 140.6B | 88.4 GB | 66K |
Community variants and fine-tunes(1)
| Model | Params | Runs from (Q4) | Context | Publisher | Quant downloads |
|---|---|---|---|---|---|
| Nous Hermes 2 Mixtral 8x7B DPO | 46.7B | 30.5 GB | 33K |
How Mixtral Compares — llmrun Composite Score
Mixtral 8x22B Instruct v0.1 is the highest-scoring Mixtral model with an llmrun Composite Score of 72.2 — #98 among 106 open models. The top proprietary model, Gemini 4 Carbon, scores 158.6. Click a model to see its full benchmark breakdown.
Gemini 4 Carbon · proprietary158.6
GPT 6 Astra · proprietary156.2
Claude Opus 5.5 · proprietary153.0
GPT 6.1 Sol · proprietary151.7
GPT 6.1 Terra · proprietary150.1
Kimi K3.1 Pro142.8
MiMo V2.6 Pro RL133.3
Kimi K3132.7
MiMo V2.6 Flash RL131.9
GLM 5.3130.1
DeepSeek V4.1 Flash129.4
DeepSeek V4 Pro 0813129.2
Frequently Asked Questions
- How much VRAM do I need to run a Mixtral model?
- The smallest Mixtral model, Tiny Mixtral, runs from 0.5 GB of VRAM at Q4_K_M (smaller low-bit quantizations start at 0.4 GB with a noticeable quality loss). Larger family members need proportionally more — see the table above for every model.
- Which Mixtral models can I run on a 16 GB GPU?
- 1 of 7 Mixtral models fit in 16 GB of VRAM at Q4_K_M, including Tiny Mixtral.
- What is the most popular Mixtral model to run locally?
- Mixtral 8x7B Instruct v0.1 is the most downloaded Mixtral model in local-friendly quantized formats. It runs from 30.5 GB of VRAM at Q4_K_M.
- How do Mixtral models score on benchmarks?
- Mixtral 8x22B Instruct v0.1 leads the family with an llmrun Composite Score of 72.2, ranking #98 among 106 open models, while the top proprietary model, Gemini 4 Carbon, scores 158.6. See the comparison chart above for the full standings.