Research2026-10-19
Meta released Muse Glimmer, an open-source language model for agent tasks. It runs on consumer hardware without cloud support. The model handles text and image inputs and outputs text. It was trained on over a hundred languages. Muse Glimmer is quantized to 4-bit for single-GPU use. It outperforms Gemma4-31B and Qwen3.6-27B on tool use and planning. The model is released under Apache 2.0. It integrates multi-step reasoning, tool use and failure recovery. The paper is available for download.
What it unlocks
Running a capable agent model locally on consumer devices without cloud infrastructure.
- 30B parameters
- 4-bit precision
- 100+ languages
What you need to act on it
- a single GPU
- Apache 2.0 license compliance
Sources