# Deploy Llama-4-Maverick-17B-128E-Instruct on dedicated GPUs

Llama-4-Maverick-17B-128E-Instruct by Meta Llama, 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/meta-llama/Llama-4-Maverick-17B-128E-Instruct.

## Specs

- Hugging Face: https://huggingface.co/meta-llama/Llama-4-Maverick-17B-128E-Instruct
- Category: Chat & reasoning
- Parameters: 402B (17B active)
- Architecture: Dense
- Context: 512K tokens
- Published precision: BF16
- Checkpoint size: 803 GB
- Minimum footprint: 2 nodes × 8 GPUs (80 GB) or 8 GPUs (141 GB) to load the weights; more for long contexts and concurrency
- License: Llama 4 Community License (Review terms)

## FAQ

### Can I run Llama-4-Maverick-17B-128E-Instruct on a dedicated deployment in Europe?

Yes. sference deploys Llama-4-Maverick-17B-128E-Instruct 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 Llama-4-Maverick-17B-128E-Instruct need?

The published checkpoint is about 803 GB (BF16), so loading it takes at least 2 nodes × 8 GPUs with 80 GB of memory each, or 8 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 Llama-4-Maverick-17B-128E-Instruct licensed for commercial use?

Llama-4-Maverick-17B-128E-Instruct is published under the Llama 4 Community License license. Meta's community license with a usage threshold, attribution requirements, and an acceptable-use policy; check its multimodal terms if your company is based in the EU.

### 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