Meta has released Muse Glimmer, a 29.6 billion parameter open-source AI model designed to run locally on devices. The model supports text and image inputs and has demonstrated superior performance compared to similarly sized models like Google's Gemma 4 31B and Qwen3.6-27B in various benchmarks. Muse Glimmer also features support for "DFlash," a speculative decoding technique that significantly speeds up processing on local hardware, such as a GPU, and is available under the Apache License 2.0. AI
IMPACT Accelerates the trend of powerful AI models becoming accessible for local, on-device execution, potentially lowering barriers for developers and users.
RANK_REASON Frontier-lab model release with system card and performance benchmarks. [lever_c_demoted from frontier_release: ic=1 ai=1.0]
Read on Mastodon — mastodon.social →
- Apache License 2.0
- Claude Haiku 4.5
- Gemma 4 31B
- Llama 4
- LM Studio
- Mark Zuckerberg
- Meta
- Muse Glimmer
- Muse Spark 1.2
- Ollama
- Qwen3.6-27B
- NVIDIA GeForce RTX 5090
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
