Qwen3.6 MoE 2-bit Quantized Version Tops Hugging Face Trending

EschaLabs · hf · 2026-07-30

The EschaLabs/Qwen3.6-35B-A3B-Escha-W2 model is trending on Hugging Face. Based on the Qwen3.5 MoE architecture, it features 35B total parameters and 3B active parameters, utilizing 2-bit quantization (EschaMoe) with support for SGLang and ZML deployment.

Original post →

More from Models

Models channel →