Skip to main content
EnterpriseModelsTools

Muse Spark Open is Meta's answer to Qwen, GLM, and Kimi in the open-weight race

Source: The Decoder (opens in a new tab) · Matthias Bastian

Intel Summary

Meta has released Muse Spark Open, an 816-billion-parameter open-weight multimodal foundation model supporting text, image, video, and audio inputs with a context window of up to one million tokens. According to reporting, the model matches leading open-weight Chinese models on benchmarks and is optimized for local execution on Apple's M5 Ultra hardware.

Why It Matters

The release provides developers and enterprises with a massive open-weight multimodal alternative featuring long-context capability. By optimizing for high-end local hardware, it challenges proprietary API providers as well as Chinese open-weight models like Qwen, GLM, and Kimi in local and private deployment scenarios.

Organizations & Entities