Qwen3.6-35B-A3B runs well on an RTX 4070, dev reports with UD-Q4_K_XL quant

haydendevs · x · 2026-09-22

Developer haydendevs shares first-hand experience that Qwen3.6-35B (MoE with 3B active parameters) runs pretty well on his RTX 4070, using the UD-Q4KXL quantization with MTP — a useful feasibility data point for running the new open-weights model on consumer GPUs.

Related event: Qwen3.6-35B Runs Smoothly on a Single RTX 4070(3 posts)→

Original post →

More from Infra

Infra channel →