QriticalBeta
Open Model Catalogue
Alibaba Qwen
Alibaba Qwen
Text generationReasoning

Qwen3.6 35B MoE

Qwen/Qwen3.6-35B-A3B

32B-class quality at 3B-active speed

Deploy Qwen3.6 35B MoE nowView on HuggingFace
Parameters
35B MoE
Min VRAM
≈ 78 GB
Hardware
1 × 80 GB GPU
OpenAI-compatible API with a per-workload key — drop-in for existing SDKs.
Dedicated GPUs — your model, your instance, no shared queues.
Billed by the second with auto-shutdown — never pay for idle time you didn't plan.