# Deploy Beam-1.0-501B-A23B on dedicated GPUs

Beam-1.0-501B-A23B by Reflection AI, 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/reflection/Beam-1.0-501B-A23B.

## Specs

- Hugging Face: https://huggingface.co/reflection/Beam-1.0-501B-A23B
- Category: Chat & reasoning
- Parameters: 501B (23B active)
- Architecture: MoE · 256 experts, 8 active
- Context: 1M tokens
- Published precision: BF16
- Checkpoint size: 1 TB
- Minimum footprint: 2 nodes × 8 GPUs (80 GB) or 2 nodes × 8 GPUs (141 GB) to load the weights; more for long contexts and concurrency
- License: Apache 2.0 (Commercial use allowed)

## FAQ

### Can I run Beam-1.0-501B-A23B on a dedicated deployment in Europe?

Yes. sference deploys Beam-1.0-501B-A23B 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 Beam-1.0-501B-A23B need?

The published checkpoint is about 1 TB (BF16), so loading it takes at least 2 nodes × 8 GPUs with 80 GB of memory each, or 2 nodes × 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 Beam-1.0-501B-A23B licensed for commercial use?

Beam-1.0-501B-A23B is published under the Apache 2.0 license. Permissive: commercial use, modification, and private deployment allowed; keep the license and notices.

### 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