~/best-gpu-for/gpt-oss-120b
OpenAI

Best GPU for gpt-oss 120B

Ranked recommendations from the 42-GPU catalog for running gpt-oss 120B (117B params). Top picks by quality, then by cost.

Top picks.

$ ./rank --by quality
Tightest budget
M5 Pro 64
64GB · runs at Q3_K_M
50 GB weights · 1.5 GB headroom
Best quality
M3 Ultra 512
512GB · runs at FP16/BF16
smallest card supporting the best available quant
Apple Silicon
M5 Pro 64
64GB · runs at Q3_K_M
smallest M-series that fits
Datacenter (FP16 ambition)
B200
192GB · runs at Q8_0
best-quality datacenter card that fits

Full ranked list.

$ ./vrambudget --rank gpt-oss-120b
GPUVRAMBest quantWeights at 15% safetyFit
appleM3 Ultra 512512GBFP16/BF16
234fitsnvidia2× H100 NVL188GBQ8_0
124fitsappleM2 Ultra 192192GBQ8_0
124fitsnvidiaB200192GBQ8_0
124fitsamdMI300X192GBQ8_0
124fitsappleM4 Max 128128GBQ6_K
96fitsappleM5 Max 128128GBQ6_K
96fitsnvidiaDGX Spark128GBQ6_K
96fitsintelGaudi 3128GBQ6_K
96fitsnvidiaH200141GBQ6_K
96fitsappleM3 Max 9696GBQ4_K_M
66fitsnvidiaRTX Pro 600096GBQ4_K_M
66fitsnvidiaH100 80GB80GBAWQ 4-BIT
62fitsappleM5 Pro 6464GBQ3_K_M
50fitsnvidiaRTX 30708GBQ4_K_M
66overnvidiaRTX 40608GBQ4_K_M
66overnvidiaRTX 308010GBQ4_K_M
66overnvidiaRTX 3060 12GB12GBQ4_K_M
66over

Next steps.

$ ./next
$ checkCan I run gpt-oss 120B?see all quants →full quant matrix$ tuneTune the mathopen calculator →/calc?model=gpt-oss-120b

Discussion.

$ gh discussion list

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