← All Models
Qwen3 30B
30B MoE (3B active) Min 19GB VRAM Full 61GB VRAM chatcodingmultilingualreasoning
Sparse Qwen3 MoE model with 256K context. Only 3B parameters are active per token, so it runs much faster than a dense 30B on 24GB-class GPUs.
Run with Ollama
ollama run qwen3:30b