// RUN ON CONSUMER GPU

Run SGPT BE 5.8B 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 → ~480 yes 188k ~$1,999
NVIDIA GeForce RTX 4090 → ~270 yes 132k ~$1,599
NVIDIA GeForce RTX 3090 Ti → ~270 yes 132k ~$1,999
NVIDIA GeForce RTX 5080 → ~257.1 yes 76k ~$999
NVIDIA GeForce RTX 3090 → ~250.7 yes 132k ~$1,499
NVIDIA GeForce RTX 3080 Ti → ~244.3 yes 48k ~$1,199
NVIDIA GeForce RTX 5070 Ti → ~240 yes 76k ~$749
AMD Radeon RX 7900 XTX → ~227.6 yes 132k ~$999
NVIDIA GeForce RTX 3080 10GB → ~203.6 yes 34k ~$699
NVIDIA GeForce RTX 4080 → ~197.1 yes 76k ~$1,199
AMD Radeon RX 7900 XT → ~189.7 yes 104k ~$899
NVIDIA GeForce RTX 5070 → ~180 yes 48k ~$549
NVIDIA GeForce RTX 4070 Ti Super → ~180 yes 76k ~$799
NVIDIA GeForce RTX 3070 Ti → ~162.9 yes 20k ~$599
AMD Radeon RX 9070 XT → ~152.7 yes 76k ~$599
AMD Radeon RX 7800 XT → ~147.9 yes 76k ~$499
NVIDIA GeForce RTX 4070 Super → ~135 yes 48k ~$599
NVIDIA GeForce RTX 4070 → ~135 yes 48k ~$599
Intel Arc Pro B70 → ~134.8 yes 188k —
Intel Arc A770 16GB → ~124.1 yes 76k ~$329
AMD Radeon RX 6800 XT → ~121.4 yes 76k ~$649
AMD Radeon RX 6900 XT → ~121.4 yes 76k ~$999
NVIDIA GeForce RTX 3070 → ~120 yes 20k ~$499
NVIDIA GeForce RTX 5060 Ti 16GB → ~120 yes 76k ~$429
NVIDIA GeForce RTX 5060 Ti 8GB → ~120 yes 20k ~$379
Intel Arc Pro B60 → ~101.1 yes 132k —
Intel Arc B580 12GB → ~101.1 yes 48k ~$249
NVIDIA GeForce RTX 4060 Ti 8GB → ~77.1 yes 20k ~$399
NVIDIA GeForce RTX 4060 Ti 16GB → ~77.1 yes 76k ~$499
NVIDIA DGX Spark → ~73.1 yes 843k ~$3,999
AMD Ryzen AI Max+ 395 → ~60.7 yes 843k ~$1,999
Intel Arc Pro B50 → ~49.7 yes 76k —

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

→ SGPT BE 5.8B overview → All hardware

FAQ

What Consumer GPU hardware runs SGPT BE 5.8B?▶

On GenAIList, 32 Consumer GPU options can run SGPT BE 5.8B at Q4 — the fastest is the NVIDIA GeForce RTX 5090 at ~480 tok/s.

How fast is SGPT BE 5.8B on Consumer GPU hardware?▶

Estimated generation speed ranges from ~49.7 to ~480 tok/s at Q4 (memory-bandwidth bound). Click any option for its full estimate and any measured numbers.