Meituan Open-Sources LongCat-Flash-Lite-Sparse Model

Meituan has open-sourced the LongCat-Flash-Lite-Sparse model, featuring an ultra-sparse MoE architecture with 69B total parameters but only 3B active. It offloads a 30B n-gram lookup table to RAM, enabling 256k context length on just 24GB of VRAM.

2026-07-31 ~ 2026-08-01 · 2 related posts

• Meituan Open-Sources LongCat-Flash-Lite-Sparse: 256k Context on a 24GB GPU — Gohab2001 · 2026-07-31
• Meituan Open-Sources LongCat-Flash-Lite: 69B Total Params with 3B Active — teortaxesTex · 2026-08-01