// RUN ON CONSUMER GPU

Run Qwen3-8B on Consumer GPU hardware

32 options ranked by estimated speed at Q4 / 8k context.

HardwareEst tok/sFits 8k?Max ctx~Price
NVIDIA GeForce RTX 5090 → ~339.5 yes 178k ~$1,999
NVIDIA GeForce RTX 4090 → ~191 yes 122k ~$1,599
NVIDIA GeForce RTX 3090 Ti → ~191 yes 122k ~$1,999
NVIDIA GeForce RTX 5080 → ~181.9 yes 66k ~$999
NVIDIA GeForce RTX 3090 → ~177.3 yes 122k ~$1,499
NVIDIA GeForce RTX 3080 Ti → ~172.8 yes 38k ~$1,199
NVIDIA GeForce RTX 5070 Ti → ~169.8 yes 66k ~$749
AMD Radeon RX 7900 XTX → ~161 yes 122k ~$999
NVIDIA GeForce RTX 3080 10GB → ~144 yes 24k ~$699
NVIDIA GeForce RTX 4080 → ~139.4 yes 66k ~$1,199
AMD Radeon RX 7900 XT → ~134.1 yes 94k ~$899
NVIDIA GeForce RTX 5070 → ~127.3 yes 38k ~$549
NVIDIA GeForce RTX 4070 Ti Super → ~127.3 yes 66k ~$799
NVIDIA GeForce RTX 3070 Ti → ~115.2 yes 10k ~$599
AMD Radeon RX 9070 XT → ~108 yes 66k ~$599
AMD Radeon RX 7800 XT → ~104.6 yes 66k ~$499
NVIDIA GeForce RTX 4070 Super → ~95.5 yes 38k ~$599
NVIDIA GeForce RTX 4070 → ~95.5 yes 38k ~$599
Intel Arc Pro B70 → ~95.3 yes 178k —
Intel Arc A770 16GB → ~87.8 yes 66k ~$329
AMD Radeon RX 6800 XT → ~85.9 yes 66k ~$649
AMD Radeon RX 6900 XT → ~85.9 yes 66k ~$999
NVIDIA GeForce RTX 3070 → ~84.9 yes 10k ~$499
NVIDIA GeForce RTX 5060 Ti 16GB → ~84.9 yes 66k ~$429
NVIDIA GeForce RTX 5060 Ti 8GB → ~84.9 yes 10k ~$379
Intel Arc Pro B60 → ~71.5 yes 122k —
Intel Arc B580 12GB → ~71.5 yes 38k ~$249
NVIDIA GeForce RTX 4060 Ti 8GB → ~54.6 yes 10k ~$399
NVIDIA GeForce RTX 4060 Ti 16GB → ~54.6 yes 66k ~$499
NVIDIA DGX Spark → ~51.7 yes 832k ~$3,999
AMD Ryzen AI Max+ 395 → ~42.9 yes 832k ~$1,999
Intel Arc Pro B50 → ~35.1 yes 66k —

Estimate at Q4 (memory-bound). Click any option for its full per-hardware estimate and measured numbers.

→ Qwen3-8B overview → All hardware

FAQ

What Consumer GPU hardware runs Qwen3-8B?▶

On GenAIList, 32 Consumer GPU options can run Qwen3-8B at Q4 — the fastest is the NVIDIA GeForce RTX 5090 at ~339.5 tok/s.

How fast is Qwen3-8B on Consumer GPU hardware?▶

Estimated generation speed ranges from ~35.1 to ~339.5 tok/s at Q4 (memory-bandwidth bound). Click any option for its full estimate and any measured numbers.