- Muse Spark Open, an 816B Mixture-of-Experts coding model with a 1M-token context window, is now available on Amazon SageMaker JumpStart for easy deployment.
- The model supports multimodal inputs (text, images, audio, video) and is optimized for long-running, multi-step coding agents with tool calling and self‑correction.
- Two deployment variants are offered: BF16 and NVFP4‑quantized, the latter using roughly a quarter of the memory for cost‑efficient inference.