~/best-gpu-for/command-r-plus
Cohere

Best GPU for Command R+

Ranked recommendations from the 42-GPU catalog for running Command R+ (104B params). Top picks by quality, then by cost.

Top picks.

$ ./rank --by quality
Tightest budget
M2 Max 64
64GB · runs at Q3_K_M
45 GB weights · 7.1 GB headroom
Best quality
M3 Ultra 512
512GB · runs at FP16/BF16
smallest card supporting the best available quant
Apple Silicon
M2 Max 64
64GB · runs at Q3_K_M
smallest M-series that fits
Datacenter (FP16 ambition)
H200
141GB · runs at Q8_0
best-quality datacenter card that fits

Full ranked list.

$ ./vrambudget --rank command-r-plus
GPUVRAMBest quantWeights at 15% safetyFit
appleM3 Ultra 512512GBFP16/BF16
208fitsnvidiaH200141GBQ8_0
111fitsnvidia2× H100 NVL188GBQ8_0
111fitsappleM2 Ultra 192192GBQ8_0
111fitsnvidiaB200192GBQ8_0
111fitsamdMI300X192GBQ8_0
111fitsappleM4 Max 128128GBQ6_K
85fitsappleM5 Max 128128GBQ6_K
85fitsnvidiaDGX Spark128GBQ6_K
85fitsintelGaudi 3128GBQ6_K
85fitsappleM3 Max 9696GBQ5_K_M
72fitsnvidiaRTX Pro 600096GBQ5_K_M
72fitsnvidiaH100 80GB80GBQ4_K_M
59fitsappleM2 Max 6464GBQ3_K_M
45fitsappleM3 Max 6464GBQ3_K_M
45fitsappleM4 Pro 6464GBQ3_K_M
45fitsappleM5 Pro 6464GBQ3_K_M
45fitsnvidiaRTX 30708GBQ4_K_M
59over

Next steps.

$ ./next
$ checkCan I run Command R+?see all quants →full quant matrix$ tuneTune the mathopen calculator →/calc?model=command-r-plus

Discussion.

$ gh discussion list

// sign in with github to leave a comment. threads live in the repo's discussions tab.