Muse Spark Open hits MLX 30 minutes after release, but even 3-bit quant needs 256GB+ RAM
pcuenq · x · 2026-10-16
- Muse Spark Open was released just 30 minutes ago and the MLX community already has a quantized version: mlx-community/Muse-Spark-Open-816B-A42B-oQ3e, runnable with mlx-lm.
- The quant uses a high-quality oQ3e format by @jundotkim, in the standard mlx-lm format so it works anywhere; the author notes it's great for agents and coding.
- It's huge: 816B parameters (A42B active) — even the 3-bit quant requires more than 256GB of RAM. Easiest way to vibe-test is via the oMLX app; follow the model card for app integration.
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