Paper page - Muse Glimmer 30B: An On-Device Agentic Model

Source: https://huggingface.co/papers/2610.24628 Authors:

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

,

Abstract

Muse Glimmer is a 30B-parameter open-source causal language model with a perception encoder that performs multi-step reasoning and tool use locally on consumer hardware via 4-bit quantization.

We introduce Muse Glimmer, an open-source 30B parametercausal language modelwithperception encoder, purpose-built for autonomous agentic tasks on consumer hardware. Muse Glimmer integratesmulti-step reasoning, reliabletool useandfailure recoveryinto a single model that runs locally on consumer devices without requiring cloud infrastructure. The model accepts text and image inputs and produces text outputs, was trained on data from over a hundred languages, and isquantizedto approximately4-bit precisionfor efficientsingle-GPU inference. Muse Glimmer achieves state-of-the-art performance for its size class onagentic benchmarks, setting a new standard compared to the models of its size: Gemma4-31B and Qwen3.6-27B ontool use, multi-step planning, and agentic task completion. We release the model under the Apache 2.0 license to foster open-source development of capable, safe, local AI agents.

View arXiv pageView PDFAdd to collection

Get this paper in your agent:

hf papers read 2610\.24628

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper3

#### meta-models/Muse-Glimmer-30B Image-Text-to-Text• 30B• Updatedabout 21 hours ago • 216k • 2.06k #### meta-models/Muse-Glimmer-30B-GGUF Image-Text-to-Text• 28B• Updatedabout 21 hours ago • 443k • 378 #### meta-models/Muse-Glimmer-30B-assistant Image-Text-to-Text• 3B• Updatedabout 21 hours ago • 101k • 75

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2610.24628 in a dataset README.md to link it from this page.

Spaces citing this paper29

Browse 29 spaces citing this paper## Collections including this paper1