// RUN ON CONSUMER GPU

Run Phi-2 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 → ~1,031.1 yes 403k ~$1,999
NVIDIA GeForce RTX 4090 → ~580 yes 291k ~$1,599
NVIDIA GeForce RTX 3090 Ti → ~580 yes 291k ~$1,999
NVIDIA GeForce RTX 5080 → ~552.4 yes 179k ~$999
NVIDIA GeForce RTX 3090 → ~538.6 yes 291k ~$1,499
NVIDIA GeForce RTX 3080 Ti → ~524.8 yes 122k ~$1,199
NVIDIA GeForce RTX 5070 Ti → ~515.6 yes 179k ~$749
AMD Radeon RX 7900 XTX → ~488.9 yes 291k ~$999
NVIDIA GeForce RTX 3080 10GB → ~437.3 yes 94k ~$699
NVIDIA GeForce RTX 4080 → ~423.5 yes 179k ~$1,199
AMD Radeon RX 7900 XT → ~407.4 yes 235k ~$899
NVIDIA GeForce RTX 5070 → ~386.7 yes 122k ~$549
NVIDIA GeForce RTX 4070 Ti Super → ~386.7 yes 179k ~$799
NVIDIA GeForce RTX 3070 Ti → ~349.8 yes 66k ~$599
AMD Radeon RX 9070 XT → ~328 yes 179k ~$599
AMD Radeon RX 7800 XT → ~317.8 yes 179k ~$499
NVIDIA GeForce RTX 4070 Super → ~290 yes 122k ~$599
NVIDIA GeForce RTX 4070 → ~290 yes 122k ~$599
Intel Arc Pro B70 → ~289.5 yes 403k —
Intel Arc A770 16GB → ~266.7 yes 179k ~$329
AMD Radeon RX 6800 XT → ~260.7 yes 179k ~$649
AMD Radeon RX 6900 XT → ~260.7 yes 179k ~$999
NVIDIA GeForce RTX 3070 → ~257.8 yes 66k ~$499
NVIDIA GeForce RTX 5060 Ti 16GB → ~257.8 yes 179k ~$429
NVIDIA GeForce RTX 5060 Ti 8GB → ~257.8 yes 66k ~$379
Intel Arc Pro B60 → ~217.1 yes 291k —
Intel Arc B580 12GB → ~217.1 yes 122k ~$249
NVIDIA GeForce RTX 4060 Ti 8GB → ~165.7 yes 66k ~$399
NVIDIA GeForce RTX 4060 Ti 16GB → ~165.7 yes 179k ~$499
NVIDIA DGX Spark → ~157.1 yes 1712k ~$3,999
AMD Ryzen AI Max+ 395 → ~130.4 yes 1712k ~$1,999
Intel Arc Pro B50 → ~106.7 yes 179k —
NVIDIA GeForce RTX 2050 Mobile → ~64.4 yes 10k —

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

→ Phi-2 overview → All hardware

FAQ

What Consumer GPU hardware runs Phi-2?▶

On GenAIList, 33 Consumer GPU options can run Phi-2 at Q4 — the fastest is the NVIDIA GeForce RTX 5090 at ~1,031.1 tok/s.

How fast is Phi-2 on Consumer GPU hardware?▶

Estimated generation speed ranges from ~64.4 to ~1,031.1 tok/s at Q4 (memory-bandwidth bound). Click any option for its full estimate and any measured numbers.