---
title: "Run Local Agentic AI Workflows with Meta’s Muse Glimmer on NVIDIA"
description: "Meta has released Muse Glimmer, a 30B open-weight dense model with a 120K+ context window aimed at local agentic AI workflows. The model uses a dense architecture that activates every parameter per token, avoiding routing overhead and targeting reliable instruction following, long-context coherence, and predictable latency for multi-step agent tasks. On NVIDIA Blackwell Ultra hardware, Muse Glimmer reports over 20K tokens/sec/GPU at BF16/NVF4 precision, with the full model fitting in a single GPU's VRAM. Deployment options span NVIDIA GeForce RTX 5090, DGX Spark, DGX Station, and Jetson edge devices, supported by vLLM, SGLang, and NVIDIA NIM containers. Fine-tuning is available through NVIDIA NeMo AutoModel for SFT and LoRA, with NeMo RL for reinforcement learning, and the NemoClaw harness enables long-running assistants in a sandboxed environment."
---

# Run Local Agentic AI Workflows with Meta’s Muse Glimmer on NVIDIA

[NVIDIA](https://yomu.fyi/company/nvidia-developer-blog) · Michelle Horton · Aug 10, 2026

**Type:** Announcement

## Summary

Meta has released Muse Glimmer, a 30B open-weight dense model with a 120K+ context window aimed at local agentic AI workflows. The model uses a dense architecture that activates every parameter per token, avoiding routing overhead and targeting reliable instruction following, long-context coherence, and predictable latency for multi-step agent tasks. On NVIDIA Blackwell Ultra hardware, Muse Glimmer reports over 20K tokens/sec/GPU at BF16/NVF4 precision, with the full model fitting in a single GPU's VRAM. Deployment options span NVIDIA GeForce RTX 5090, DGX Spark, DGX Station, and Jetson edge devices, supported by vLLM, SGLang, and NVIDIA NIM containers. Fine-tuning is available through NVIDIA NeMo AutoModel for SFT and LoRA, with NeMo RL for reinforcement learning, and the NemoClaw harness enables long-running assistants in a sandboxed environment.

## Context

Most LLMs are optimized for chat and single-turn interactions, but agentic workloads such as software scaffolding, documentation revision, and knowledge base management require sustained throughput, long-context coherence, and reliability across sequential tool calls. Privacy and compliance needs also push inference onto local hardware.

## Approach / What changed

Meta released Muse Glimmer, a 30B dense open-weight model with a 120K+ context window that activates every parameter per token. It is optimized for NVIDIA Tensor Core platforms including GeForce RTX 5090, DGX Spark, DGX Station, and Jetson, and is deployable via vLLM, SGLang, or NVIDIA NIM containers, with fine-tuning through NeMo AutoModel and NeMo RL.

## Takeaways

- Muse Glimmer is a 30B dense open-weight model with a 120K+ context window, designed for agentic workloads rather than chat, and avoids MoE routing overhead by activating every parameter per token.
- On NVIDIA Blackwell Ultra, the model delivers over 20K tokens/sec/GPU at BF16/NVF4 precision and fits entirely in a single GPU's VRAM, enabling local inference without sharding or CPU offloading.
- Deployment spans GeForce RTX 5090, DGX Spark, DGX Station, and Jetson edge devices using vLLM, SGLang, or NVIDIA NIM containers, with fine-tuning support via NeMo AutoModel (SFT and LoRA) and NeMo RL, plus the NemoClaw harness for long-running assistants.

**Tags:** [Deployment](https://yomu.fyi/topic/deployment), [LLMs](https://yomu.fyi/topic/llm), [Open Source](https://yomu.fyi/topic/open-source), [Performance](https://yomu.fyi/topic/performance)

- Source: [NVIDIA](https://developer.nvidia.com/blog/run-local-agentic-ai-workflows-with-metas-muse-glimmer-on-nvidia)
- Source URL: https://developer.nvidia.com/blog/run-local-agentic-ai-workflows-with-metas-muse-glimmer-on-nvidia
- Ingested by Yomu: 2026-08-28T08:00:47.427Z

[Read original post](https://developer.nvidia.com/blog/run-local-agentic-ai-workflows-with-metas-muse-glimmer-on-nvidia)