---
type: "Article"
title: "Beam-1.0-501B-A23B EXL3 3 bpw with vLLM on 2× DGX Spark"
description: "Beam-1.0-501B-A23B EXL3 3 bpw with vLLM on 2× DGX Spark: recipe by bertholomus with speed, condition and source for every number, weights and license."
resource: "https://www.contextstudios.ai/local-ai/bertholomus--beam-1.0-501b-a23b-exl3-3bpw-vllm-2x-dgx-spark"
language: "en"
generated:
  by: "process:contextstudios-md/1"
  at: "2026-11-11T03:30:22.210Z"
status: "stable"
---

# Beam-1.0-501B-A23B EXL3 3 bpw with vLLM on 2× DGX Spark

Beam-1.0-501B-A23B EXL3 3 bpw with vLLM on 2× DGX Spark: 28.3 tok/s according to github.com (dataset as of Nov 3, 2026).

- Creator: [bertholomus](https://www.contextstudios.ai/local-ai/creators/bertholomus.md)
- Engine: vLLM TP2 (image beam-exl3-gb10 via gb10/build-image.sh: Reflection arm64 Beam vLLM image + vllm-exl3 d3cfd394 with Diffbot Beam patch, ExLlamaV3 5be88657 for sm_121, kernels xmoe_dense/xmoe_do8/x8rh/f8rh), DFlash2 drafter (3 draft tokens)
- Quantization: EXL3 3 bpw (routed experts 3 bit, attention/dense MLPs 6 bit, embedding/norms/router/lm_head BF16; ~90 GiB per node)
- Model family: Beam-1.0-501B-A23B
- Hardware: dgx-spark x2
- Context: 655360

## All numbers

- 28.3 tok/s (Everyday; Everyday, 1 request, prose prompt (~2K), with DFlash2 (per recipe)) — [source](https://github.com/bertholomus/beam-1.0-2x-dgx-spark#measured)
- decode_tps: 28.3 (1 request) — [source](https://github.com/bertholomus/beam-1.0-2x-dgx-spark#measured)
- decode_tps: 33.4 (1 request) — [source](https://github.com/bertholomus/beam-1.0-2x-dgx-spark#measured)
- prefill_tps: 1362 (1 request) — [source](https://github.com/bertholomus/beam-1.0-2x-dgx-spark#measured)
- ttft_ms: 440 (1 request) — [source](https://github.com/bertholomus/beam-1.0-2x-dgx-spark#measured)

## Weights

- [bertholomus/Beam-1.0-501B-A23B-EXL3-3bpw-2xGB10](https://huggingface.co/bertholomus/Beam-1.0-501B-A23B-EXL3-3bpw-2xGB10) (main, apache-2.0)
- [reflection/Beam-1.0-501B-A23B-DFlash2](https://huggingface.co/reflection/Beam-1.0-501B-A23B-DFlash2) (draft, apache-2.0)
- [reflection/Beam-1.0-501B-A23B](https://huggingface.co/reflection/Beam-1.0-501B-A23B) (base, apache-2.0)

[https://github.com/bertholomus/beam-1.0-2x-dgx-spark](https://github.com/bertholomus/beam-1.0-2x-dgx-spark)

## Related

- [Local AI on your hardware.](https://www.contextstudios.ai/local-ai.md)