// RUN ON CONSUMER GPU

Run Pythia-160m 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 → ~17,400 yes 425k ~$1,999
NVIDIA GeForce RTX 4090 → ~9,787.5 yes 313k ~$1,599
NVIDIA GeForce RTX 3090 Ti → ~9,787.5 yes 313k ~$1,999
NVIDIA GeForce RTX 5080 → ~9,321.4 yes 200k ~$999
NVIDIA GeForce RTX 3090 → ~9,088.4 yes 313k ~$1,499
NVIDIA GeForce RTX 3080 Ti → ~8,855.4 yes 144k ~$1,199
NVIDIA GeForce RTX 5070 Ti → ~8,700 yes 200k ~$749
AMD Radeon RX 7900 XTX → ~8,250 yes 313k ~$999
NVIDIA GeForce RTX 3080 10GB → ~7,379.5 yes 116k ~$699
NVIDIA GeForce RTX 4080 → ~7,146.4 yes 200k ~$1,199
AMD Radeon RX 7900 XT → ~6,875 yes 257k ~$899
NVIDIA GeForce RTX 5070 → ~6,525 yes 144k ~$549
NVIDIA GeForce RTX 4070 Ti Super → ~6,525 yes 200k ~$799
NVIDIA GeForce RTX 3070 Ti → ~5,903.6 yes 88k ~$599
AMD Radeon RX 9070 XT → ~5,534.4 yes 200k ~$599
AMD Radeon RX 7800 XT → ~5,362.5 yes 200k ~$499
NVIDIA GeForce RTX 4070 Super → ~4,893.8 yes 144k ~$599
NVIDIA GeForce RTX 4070 → ~4,893.8 yes 144k ~$599
Intel Arc Pro B70 → ~4,885.7 yes 425k —
Intel Arc A770 16GB → ~4,500 yes 200k ~$329
AMD Radeon RX 6800 XT → ~4,400 yes 200k ~$649
AMD Radeon RX 6900 XT → ~4,400 yes 200k ~$999
NVIDIA GeForce RTX 3070 → ~4,350 yes 88k ~$499
NVIDIA GeForce RTX 5060 Ti 16GB → ~4,350 yes 200k ~$429
NVIDIA GeForce RTX 5060 Ti 8GB → ~4,350 yes 88k ~$379
Intel Arc Pro B60 → ~3,664.3 yes 313k —
Intel Arc B580 12GB → ~3,664.3 yes 144k ~$249
NVIDIA GeForce RTX 4060 Ti 8GB → ~2,796.4 yes 88k ~$399
NVIDIA GeForce RTX 4060 Ti 16GB → ~2,796.4 yes 200k ~$499
NVIDIA DGX Spark → ~2,650.8 yes 1734k ~$3,999
AMD Ryzen AI Max+ 395 → ~2,200 yes 1734k ~$1,999
Intel Arc Pro B50 → ~1,800 yes 200k —
NVIDIA GeForce RTX 2050 Mobile → ~1,087.5 yes 32k —

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

→ Pythia-160m overview → All hardware

FAQ

What Consumer GPU hardware runs Pythia-160m?▶

On GenAIList, 33 Consumer GPU options can run Pythia-160m at Q4 — the fastest is the NVIDIA GeForce RTX 5090 at ~17,400 tok/s.

How fast is Pythia-160m on Consumer GPU hardware?▶

Estimated generation speed ranges from ~1,087.5 to ~17,400 tok/s at Q4 (memory-bandwidth bound). Click any option for its full estimate and any measured numbers.