// RUN ON CONSUMER GPU

Run RWKV-6 (Finch) 3B on Consumer GPU hardware

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

HardwareEst tok/sFits 8k?Max ctx~Price
NVIDIA GeForce RTX 5090 → ~898.1 yes 400k ~$1,999
NVIDIA GeForce RTX 4090 → ~505.2 yes 288k ~$1,599
NVIDIA GeForce RTX 3090 Ti → ~505.2 yes 288k ~$1,999
NVIDIA GeForce RTX 5080 → ~481.1 yes 175k ~$999
NVIDIA GeForce RTX 3090 → ~469.1 yes 288k ~$1,499
NVIDIA GeForce RTX 3080 Ti → ~457.1 yes 119k ~$1,199
NVIDIA GeForce RTX 5070 Ti → ~449 yes 175k ~$749
AMD Radeon RX 7900 XTX → ~425.8 yes 288k ~$999
NVIDIA GeForce RTX 3080 10GB → ~380.9 yes 91k ~$699
NVIDIA GeForce RTX 4080 → ~368.8 yes 175k ~$1,199
AMD Radeon RX 7900 XT → ~354.8 yes 231k ~$899
NVIDIA GeForce RTX 5070 → ~336.8 yes 119k ~$549
NVIDIA GeForce RTX 4070 Ti Super → ~336.8 yes 175k ~$799
NVIDIA GeForce RTX 3070 Ti → ~304.7 yes 63k ~$599
AMD Radeon RX 9070 XT → ~285.6 yes 175k ~$599
AMD Radeon RX 7800 XT → ~276.8 yes 175k ~$499
NVIDIA GeForce RTX 4070 Super → ~252.6 yes 119k ~$599
NVIDIA GeForce RTX 4070 → ~252.6 yes 119k ~$599
Intel Arc Pro B70 → ~252.2 yes 400k —
Intel Arc A770 16GB → ~232.3 yes 175k ~$329
AMD Radeon RX 6800 XT → ~227.1 yes 175k ~$649
AMD Radeon RX 6900 XT → ~227.1 yes 175k ~$999
NVIDIA GeForce RTX 3070 → ~224.5 yes 63k ~$499
NVIDIA GeForce RTX 5060 Ti 16GB → ~224.5 yes 175k ~$429
NVIDIA GeForce RTX 5060 Ti 8GB → ~224.5 yes 63k ~$379
Intel Arc Pro B60 → ~189.1 yes 288k —
Intel Arc B580 12GB → ~189.1 yes 119k ~$249
NVIDIA GeForce RTX 4060 Ti 8GB → ~144.3 yes 63k ~$399
NVIDIA GeForce RTX 4060 Ti 16GB → ~144.3 yes 175k ~$499
NVIDIA DGX Spark → ~136.8 yes 1708k ~$3,999
AMD Ryzen AI Max+ 395 → ~113.5 yes 1708k ~$1,999
Intel Arc Pro B50 → ~92.9 yes 175k —
NVIDIA GeForce RTX 2050 Mobile → ~56.1 tight 7k —

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

→ RWKV-6 (Finch) 3B overview → All hardware

FAQ

What Consumer GPU hardware runs RWKV-6 (Finch) 3B?▶

On GenAIList, 33 Consumer GPU options can run RWKV-6 (Finch) 3B at Q4 — the fastest is the NVIDIA GeForce RTX 5090 at ~898.1 tok/s.

How fast is RWKV-6 (Finch) 3B on Consumer GPU hardware?▶

Estimated generation speed ranges from ~56.1 to ~898.1 tok/s at Q4 (memory-bandwidth bound). Click any option for its full estimate and any measured numbers.