# Deploy MiniMax-M3.1-Flash on dedicated GPUs

MiniMax-M3.1-Flash by MiniMax, deployable on sference as a dedicated deployment: reserved GPU capacity, private endpoint, European GPUs, billed per GPU-hour on monthly commitments. Request it at https://sference.com/models/MiniMaxAI/MiniMax-M3.1-Flash.

## Specs

- Hugging Face: https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash
- Category: Chat & reasoning
- Parameters: 220B
- Architecture: MoE · 128 experts, 4 active
- Context: 1M tokens
- Published precision: 4-bit (packed) + BF16
- Checkpoint size: 250 GB
- Minimum footprint: 4 GPUs (80 GB) or 4 GPUs (141 GB) to load the weights; more for long contexts and concurrency
- License: Custom license (Review terms) — LICENSE

## FAQ

### Can I run MiniMax-M3.1-Flash on a dedicated deployment in Europe?

Yes. sference deploys MiniMax-M3.1-Flash on GPU capacity reserved for you, behind a private endpoint, with every request processed on European GPUs and logged in your audit trail. Tell us your expected volume and latency targets with the form on this page and we come back with a sizing and a quote.

### How many GPUs does MiniMax-M3.1-Flash need?

The published checkpoint is about 250 GB (4-bit (packed) + BF16), so loading it takes at least 4 GPUs with 80 GB of memory each, or 4 GPUs with 141 GB each. Long contexts and high concurrency need more for the KV cache; quantization needs less. You don't pick hardware on sference: you choose a performance tier and we size the deployment.

### Is MiniMax-M3.1-Flash licensed for commercial use?

MiniMax-M3.1-Flash is published under the Custom license license. Custom license from the model maker (minimax-community). Review the terms on the model page before production use.

### How is a dedicated deployment priced?

Per GPU-hour on a monthly commitment, sized to your throughput and latency targets.

Facts from Hugging Face, synced 2026-10-27. All deployable models: https://sference.com/dedicated.md