// RUN ON CONSUMER GPU

Run Pythia-6.9b 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 → ~403.5 yes 184k ~$1,999
NVIDIA GeForce RTX 4090 → ~227 yes 128k ~$1,599
NVIDIA GeForce RTX 3090 Ti → ~227 yes 128k ~$1,999
NVIDIA GeForce RTX 5080 → ~216.1 yes 71k ~$999
NVIDIA GeForce RTX 3090 → ~210.7 yes 128k ~$1,499
NVIDIA GeForce RTX 3080 Ti → ~205.3 yes 43k ~$1,199
NVIDIA GeForce RTX 5070 Ti → ~201.7 yes 71k ~$749
AMD Radeon RX 7900 XTX → ~191.3 yes 128k ~$999
NVIDIA GeForce RTX 3080 10GB → ~171.1 yes 29k ~$699
NVIDIA GeForce RTX 4080 → ~165.7 yes 71k ~$1,199
AMD Radeon RX 7900 XT → ~159.4 yes 99k ~$899
NVIDIA GeForce RTX 5070 → ~151.3 yes 43k ~$549
NVIDIA GeForce RTX 4070 Ti Super → ~151.3 yes 71k ~$799
NVIDIA GeForce RTX 3070 Ti → ~136.9 yes 15k ~$599
AMD Radeon RX 9070 XT → ~128.3 yes 71k ~$599
AMD Radeon RX 7800 XT → ~124.3 yes 71k ~$499
NVIDIA GeForce RTX 4070 Super → ~113.5 yes 43k ~$599
NVIDIA GeForce RTX 4070 → ~113.5 yes 43k ~$599
Intel Arc Pro B70 → ~113.3 yes 184k —
Intel Arc A770 16GB → ~104.3 yes 71k ~$329
AMD Radeon RX 6800 XT → ~102 yes 71k ~$649
AMD Radeon RX 6900 XT → ~102 yes 71k ~$999
NVIDIA GeForce RTX 3070 → ~100.9 yes 15k ~$499
NVIDIA GeForce RTX 5060 Ti 16GB → ~100.9 yes 71k ~$429
NVIDIA GeForce RTX 5060 Ti 8GB → ~100.9 yes 15k ~$379
Intel Arc Pro B60 → ~85 yes 128k —
Intel Arc B580 12GB → ~85 yes 43k ~$249
NVIDIA GeForce RTX 4060 Ti 8GB → ~64.8 yes 15k ~$399
NVIDIA GeForce RTX 4060 Ti 16GB → ~64.8 yes 71k ~$499
NVIDIA DGX Spark → ~61.5 yes 838k ~$3,999
AMD Ryzen AI Max+ 395 → ~51 yes 838k ~$1,999
Intel Arc Pro B50 → ~41.7 yes 71k —

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

→ Pythia-6.9b overview → All hardware

FAQ

What Consumer GPU hardware runs Pythia-6.9b?▶

On GenAIList, 32 Consumer GPU options can run Pythia-6.9b at Q4 — the fastest is the NVIDIA GeForce RTX 5090 at ~403.5 tok/s.

How fast is Pythia-6.9b on Consumer GPU hardware?▶

Estimated generation speed ranges from ~41.7 to ~403.5 tok/s at Q4 (memory-bandwidth bound). Click any option for its full estimate and any measured numbers.