Meta releases Muse Spark Open: 816B/42B open model hits 190 tok/s on B300 via TokenSpeed
PyTorch · x · 2026-10-16
Meta has released Muse Spark Open, billed as the most capable US open-weight model to date, built for coding and long-horizon agentic work.
Key specs:
- 816B total / 42B active parameters (MoE)
- 1M token context window
- Natively multimodal across text, image, video, and audio
On the inference side, LightSeek's open-source TokenSpeed engine shipped day-0 support: 190 output tokens/s per user on NVIDIA B300 (TP4/EP4, 8K in / 1K out) without speculative decoding, with AMD Instinct MI355X also supported. TokenSpeed has 2.4k stars on GitHub and is MIT-licensed.
Related event: Meta Open-Sources Muse Spark Open, Its Strongest Open-Weight Model(59 posts)→
More from Infra
- FT warns of 'Darwin moment' for AI borrowers as CoreWeave bonds sink to distress levels — firstadopter · 2026-10-16
- Phone as interface, home servers as the lab: a two-workflow local AI setup — EAccelerate_42 · 2026-10-16
- "A datacenter is a bank": CoreWeave drama revives calls for算力 supervision — arthurcolle · 2026-10-16
- DOT-MoE: converting dense models into sparse MoE via optimal transport — burkov · 2026-10-16
- Stanford CS336: A 7B model needs 14GB for weights but 84GB to train — stanfordnlp · 2026-10-16
- New NVFP4 trick: compressing block scales saves 45-48% per expert matrix, bit-exact — TheZachMueller · 2026-10-16