// RUN ON CONSUMER GPU

Run Adaptive Inputs + LayerDrop 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 → ~6,581.6 yes 423k ~$1,999
NVIDIA GeForce RTX 4090 → ~3,702.1 yes 310k ~$1,599
NVIDIA GeForce RTX 3090 Ti → ~3,702.1 yes 310k ~$1,999
NVIDIA GeForce RTX 5080 → ~3,525.8 yes 198k ~$999
NVIDIA GeForce RTX 3090 → ~3,437.7 yes 310k ~$1,499
NVIDIA GeForce RTX 3080 Ti → ~3,349.5 yes 142k ~$1,199
NVIDIA GeForce RTX 5070 Ti → ~3,290.8 yes 198k ~$749
AMD Radeon RX 7900 XTX → ~3,120.6 yes 310k ~$999
NVIDIA GeForce RTX 3080 10GB → ~2,791.3 yes 114k ~$699
NVIDIA GeForce RTX 4080 → ~2,703.1 yes 198k ~$1,199
AMD Radeon RX 7900 XT → ~2,600.5 yes 254k ~$899
NVIDIA GeForce RTX 5070 → ~2,468.1 yes 142k ~$549
NVIDIA GeForce RTX 4070 Ti Super → ~2,468.1 yes 198k ~$799
NVIDIA GeForce RTX 3070 Ti → ~2,233 yes 86k ~$599
AMD Radeon RX 9070 XT → ~2,093.4 yes 198k ~$599
AMD Radeon RX 7800 XT → ~2,028.4 yes 198k ~$499
NVIDIA GeForce RTX 4070 Super → ~1,851.1 yes 142k ~$599
NVIDIA GeForce RTX 4070 → ~1,851.1 yes 142k ~$599
Intel Arc Pro B70 → ~1,848 yes 423k —
Intel Arc A770 16GB → ~1,702.1 yes 198k ~$329
AMD Radeon RX 6800 XT → ~1,664.3 yes 198k ~$649
AMD Radeon RX 6900 XT → ~1,664.3 yes 198k ~$999
NVIDIA GeForce RTX 3070 → ~1,645.4 yes 86k ~$499
NVIDIA GeForce RTX 5060 Ti 16GB → ~1,645.4 yes 198k ~$429
NVIDIA GeForce RTX 5060 Ti 8GB → ~1,645.4 yes 86k ~$379
Intel Arc Pro B60 → ~1,386 yes 310k —
Intel Arc B580 12GB → ~1,386 yes 142k ~$249
NVIDIA GeForce RTX 4060 Ti 8GB → ~1,057.8 yes 86k ~$399
NVIDIA GeForce RTX 4060 Ti 16GB → ~1,057.8 yes 198k ~$499
NVIDIA DGX Spark → ~1,002.7 yes 1731k ~$3,999
AMD Ryzen AI Max+ 395 → ~832.2 yes 1731k ~$1,999
Intel Arc Pro B50 → ~680.9 yes 198k —
NVIDIA GeForce RTX 2050 Mobile → ~411.3 yes 30k —

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

→ Adaptive Inputs + LayerDrop overview → All hardware

FAQ

What Consumer GPU hardware runs Adaptive Inputs + LayerDrop?▶

On GenAIList, 33 Consumer GPU options can run Adaptive Inputs + LayerDrop at Q4 — the fastest is the NVIDIA GeForce RTX 5090 at ~6,581.6 tok/s.

How fast is Adaptive Inputs + LayerDrop on Consumer GPU hardware?▶

Estimated generation speed ranges from ~411.3 to ~6,581.6 tok/s at Q4 (memory-bandwidth bound). Click any option for its full estimate and any measured numbers.