New Delhi [India], August 11 : Meta has opened the weights of its 30-billion-parameter Muse Glimmer artificial intelligence model, while CEO Mark Zuckerberg said the company will soon release the weights of Muse Spark 1.2, its latest foundation model.
"Today we're also opening the weights for Muse Glimmer, a great 30B parameter dense model that can run locally. Soon we'll also release the weights for Muse Spark 1.2, our latest foundation model," Zuckerberg said in a post on Monday.
Zuckerberg said Meta is a "strong supporter of open source" and expressed pride in the releases.
The announcement marks a further step in Meta's release of model weights, with the company making Muse Glimmer available to developers for use in building and customising AI applications.
Muse Glimmer, developed by Meta Superintelligence Labs, is a 30-billion-parameter dense model designed to run locally on consumer computers. Meta had said the model is designed for "always-on local agent workflows" and can run on a Mac or PC equipped with a single consumer graphics processing unit (GPU).
The model is aimed at AI agents that can carry out tasks on behalf of users, including managing schedules, drafting messages and organising files. It can also be used for coding and other developer applications.
Muse Glimmer can process both text and images and has been trained on data from more than 100 languages, according to Meta.
Meta said it compressed the model to under 20 GB using quantisation, allowing the model and its supporting components to operate within a 24 GB or 32 GB memory envelope. The company tested it on MacBook M4 Max and M5 Max devices as well as NVIDIA's RTX 5090 GPU.
Meta said Muse Glimmer is available for developers to download on Hugging Face, with integrations with platforms and frameworks including Ollama, LM Studio, llama.cpp, ExecuTorch and MLX expected.
The company is also working with AMD, Arm, Dell, Intel and NVIDIA to further optimise the model's performance across devices.
Zuckerberg's announcement of Muse Spark 1.2's upcoming weight release points to another model release from Meta in its ongoing push to make its AI models more accessible to developers.