The download is available for free and users can run the template on their own PC.
Meta has released a new lightweight “open source” AI model that’s light enough to run on a single computer, the company announced today. Called Muse Glimmer, it’s based on Meta’s Spark 1.2 closed model, but is small enough to require a single GPU for agent-oriented tasks like scheduling and file management. “We designed Muse Glimmer to balance capabilities with the memory and compute constraints of local hardware,” the company wrote.
Meta said it was making the “weights” that AI systems use to choose answers available to everyone on Hugging Face, along with developer documentation. The download is available for free and users can run the template on their own PC. The company said optimized integrations will be available on llama.cpp and other sites, “so you can go from download to working agent in minutes.”
The model is powerful for its size, according to Meta, with “high success rates” on benchmarks like DeepSearch QA, MCP-Atlas, and SWE-Bench (which evaluates its ability to write and debug code). It also supports reliable tool usage, multi-step reasoning, crash recovery, multimodal input, and scaffolding compatibility for working with OpenClaw and other agent orchestrators. It was trained on data from more than 100 languages, the company added.
Meta’s new model is reminiscent of some Chinese AI competitors like DeepSeek, which can run on local machines and offers more open licensing. “Rather than centralizing superintelligence, we should disseminate it widely and give everyone the opportunity to direct it,” CEO Mark Zuckerberg said in an essay accompanying the release of Muse Glimmer.
Meta’s Muse Spark AI is considered weaker than its rivals OpenAI and Anthtropic. So the new model looks like an effort to take its AI in a more open source direction and attract users who prefer to run their models locally. This has been a hot topic lately, with NVIDIA recently launching the Open Secure AI Alliance to improve cyber defense following the malicious attack on Hugging Face by a previously unreleased, closed OpenAI model.
