Tok/s vs cost of self-hosting step3p7-flash on H200:8.
Run it yourself
step-3.7-flash@v2 · H200:8
$ iwant launch step-3.7-flashResults
Dashed lines estimate the same throughput at another cloud's price for the same GPUs.
| setup | $/hour | peak tok/s | $/1M at peak | |
|---|---|---|---|---|
| step-3.7-flash@v2 · GCP a3-ultragpu-8g H200:8 us-west1-c Spot | $52.03 | 12,437 | $1.1620 | |
| step-3.7-flash@v2 · AWS ap-northeast-2 spot (estimated) | $12.18 | 12,437 | $0.2721 | |
| step-3.7-flash@v2 · AWS us-east-1 on-demand (estimated) | $63.30 | 12,437 | $1.4137 |