// RUN ON CONSUMER GPU

Run Codestral Mamba 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 → ~382.1 yes 182k ~$1,999
NVIDIA GeForce RTX 4090 → ~215 yes 126k ~$1,599
NVIDIA GeForce RTX 3090 Ti → ~215 yes 126k ~$1,999
NVIDIA GeForce RTX 5080 → ~204.7 yes 70k ~$999
NVIDIA GeForce RTX 3090 → ~199.6 yes 126k ~$1,499
NVIDIA GeForce RTX 3080 Ti → ~194.5 yes 42k ~$1,199
NVIDIA GeForce RTX 5070 Ti → ~191.1 yes 70k ~$749
AMD Radeon RX 7900 XTX → ~181.2 yes 126k ~$999
NVIDIA GeForce RTX 3080 10GB → ~162.1 yes 28k ~$699
NVIDIA GeForce RTX 4080 → ~156.9 yes 70k ~$1,199
AMD Radeon RX 7900 XT → ~151 yes 98k ~$899
NVIDIA GeForce RTX 5070 → ~143.3 yes 42k ~$549
NVIDIA GeForce RTX 4070 Ti Super → ~143.3 yes 70k ~$799
NVIDIA GeForce RTX 3070 Ti → ~129.7 yes 14k ~$599
AMD Radeon RX 9070 XT → ~121.5 yes 70k ~$599
AMD Radeon RX 7800 XT → ~117.8 yes 70k ~$499
NVIDIA GeForce RTX 4070 Super → ~107.5 yes 42k ~$599
NVIDIA GeForce RTX 4070 → ~107.5 yes 42k ~$599
Intel Arc Pro B70 → ~107.3 yes 182k —
Intel Arc A770 16GB → ~98.8 yes 70k ~$329
AMD Radeon RX 6800 XT → ~96.6 yes 70k ~$649
AMD Radeon RX 6900 XT → ~96.6 yes 70k ~$999
NVIDIA GeForce RTX 3070 → ~95.5 yes 14k ~$499
NVIDIA GeForce RTX 5060 Ti 16GB → ~95.5 yes 70k ~$429
NVIDIA GeForce RTX 5060 Ti 8GB → ~95.5 yes 14k ~$379
Intel Arc Pro B60 → ~80.5 yes 126k —
Intel Arc B580 12GB → ~80.5 yes 42k ~$249
NVIDIA GeForce RTX 4060 Ti 8GB → ~61.4 yes 14k ~$399
NVIDIA GeForce RTX 4060 Ti 16GB → ~61.4 yes 70k ~$499
NVIDIA DGX Spark → ~58.2 yes 836k ~$3,999
AMD Ryzen AI Max+ 395 → ~48.3 yes 836k ~$1,999
Intel Arc Pro B50 → ~39.5 yes 70k —

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

→ Codestral Mamba overview → All hardware

FAQ

What Consumer GPU hardware runs Codestral Mamba?▶

On GenAIList, 32 Consumer GPU options can run Codestral Mamba at Q4 — the fastest is the NVIDIA GeForce RTX 5090 at ~382.1 tok/s.

How fast is Codestral Mamba on Consumer GPU hardware?▶

Estimated generation speed ranges from ~39.5 to ~382.1 tok/s at Q4 (memory-bandwidth bound). Click any option for its full estimate and any measured numbers.