Qwen3 30B-A3B Instruct 2507 · 2× NVIDIA GeForce RTX 4090 · fp8

$/hr: $0.70
Weights / VRAM: 31 / 48 GB
FP16 TFLOPS: 330
Params: 30.5B · qwen3_moe
Context: 262,144
License: apache-2.0

← all offerings · deployments