---
type: "post"
id: "scale-agentic-coding-workflows-with-meta-s-muse-8364b4"
slug: "scale-agentic-coding-workflows-with-meta-s-muse-8364b4"
title: "Scale Agentic Coding Workflows with Meta's Muse Spark Open on NVIDIA GB300 NVL72"
description: "NVIDIA has published deployment guidance for Meta's Muse Spark Open on its GB300 NVL72 Blackwell Ultra platform."
retrieval_nugget: "NVIDIA has published deployment guidance for Meta's Muse Spark Open on its GB300 NVL72 Blackwell Ultra platform."
published_at: "2026-10-23"
updated_at: "2026-10-23"
record_date: "2026-10-23"
date_kind: "published_at"
topics: ["coding-agents","agent-workflows","compute"]
entities: ["NVIDIA","Meta","Muse Spark Open","NVIDIA GB300 NVL72","NVIDIA Dynamo","NVIDIA NIM","TensorRT LLM","vLLM"]
related_patterns: []
novelty_label: "new_fact"
what_changed: "NVIDIA details Muse Spark Open deployment on GB300 NVL72, including NVFP4 optimization and throughput claims of up to 50K tokens/s per GPU and 600 tokens/s per user."
prior_refs: ["memory:9b5c107c-245f-8900-aff0-653c2698f524","memory:24ae215f-5331-84d3-a726-5e81580f2590"]
source_url: "https://developer.nvidia.com/blog/scale-agentic-coding-workflows-with-metas-muse-spark-open-on-nvidia-gb300-nvl72"
source_urls: ["https://developer.nvidia.com/blog/scale-agentic-coding-workflows-with-metas-muse-spark-open-on-nvidia-gb300-nvl72"]
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "post:scale-agentic-coding-workflows-with-meta-s-muse-8364b4"
status: "published"
visuals: []
editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]}
routes: {"html":"https://newruntime.com/posts/scale-agentic-coding-workflows-with-meta-s-muse-8364b4/","markdown":"https://newruntime.com/posts/scale-agentic-coding-workflows-with-meta-s-muse-8364b4.md","json":"https://newruntime.com/posts/scale-agentic-coding-workflows-with-meta-s-muse-8364b4.json"}
---

# Scale Agentic Coding Workflows with Meta's Muse Spark Open on NVIDIA GB300 NVL72

## Retrieval answer

NVIDIA has published deployment guidance for Meta's Muse Spark Open on its GB300 NVL72 Blackwell Ultra platform.

NVIDIA has published deployment guidance for Meta's Muse Spark Open on its GB300 NVL72 Blackwell Ultra platform. The post describes Muse Spark Open as a mixture-of-experts coding model with 816 billion total parameters, 42 billion active parameters per token, and a one-million-token context window.

NVIDIA says NVFP4 checkpoints are optimized for Blackwell Ultra, and points developers to SGLang, vLLM, TensorRT LLM, and NVIDIA Dynamo deployment options. It also says the weights can be downloaded from Hugging Face and that a downloadable NVIDIA NIM is available for production setup. NVIDIA describes GB300 NVL72 as a 72-GPU NVLink domain.

For agentic workloads, NVIDIA claims up to 50,000 tokens per second per GPU and up to 600 tokens per second per user on four Blackwell GPUs with speculative decoding. The source does not establish an independently reproduced benchmark, the exact workload and concurrency behind every result, comparative performance on other hardware, or deployment pricing.

Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json

## Continue in the graph

Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`.

- topic hub: [Coding Agents](https://newruntime.com/topics/coding-agents.md): Explore the coding-agents topic hub.
- topic hub: [Agent Workflows](https://newruntime.com/topics/agent-workflows.md): Explore the agent-workflows topic hub.
- topic hub: [Compute](https://newruntime.com/topics/compute.md): Explore the compute topic hub.
- collection: [Field Notes](https://newruntime.com/posts.json): Open the latest editorial analysis.
- collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph.
- whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)