inference/research
verifiedhigh · 0.95updated 2026-10-04

Replicate Pricing - Per-second hardware rates for CPU and GPU instances

database

Fields

id
sources/ppl-cloud-replicate-pricing
Publisher
Replicate
Kind
database
Accessed
2026-10-04
notes
Fetched 2026-10-04. Per-second billing, shown both per-second and per-hour. Hardware tiers: CPU Small $0.000025/sec ($0.09/hr); CPU $0.000100/sec ($0.36/hr); T4 $0.000225/sec ($0.81/hr); L40S $0.000975/sec ($3.51/hr); 2x L40S $0.001950/sec ($7.02/hr); A100 80GB $0.001400/sec ($5.04/hr); 2x A100 $0.002800/sec ($10.08/hr); H100 $0.001525/sec ($5.49/hr); 2x H100 $0.003050/sec ($10.98/hr); 4x H100 $0.006100/sec ($21.96/hr); 8x H100 $0.012200/sec ($43.92/hr); H200 $0.001525/sec ($5.49/hr); 2x H200 $0.003050/sec ($10.98/hr); 4x H200 $0.006100/sec ($21.96/hr); 8x H200 $0.012200/sec ($43.92/hr); 4x A100 $0.005600/sec ($20.16/hr); 8x A100 $0.011200/sec ($40.32/hr); 4x L40S $0.003900/sec ($14.04/hr); 8x L40S $0.007800/sec ($28.08/hr). Multi-GPU tiers above 1x require committed spend contracts. Public models billed by time or per-input/output; private models on dedicated hardware billed for all time online (including idle). Enterprise & volume discounts available.

Notes

Fetched 2026-10-04. Per-second billing, shown both per-second and per-hour. Hardware tiers: CPU Small $0.000025/sec ($0.09/hr); CPU $0.000100/sec ($0.36/hr); T4 $0.000225/sec ($0.81/hr); L40S $0.000975/sec ($3.51/hr); 2x L40S $0.001950/sec ($7.02/hr); A100 80GB $0.001400/sec ($5.04/hr); 2x A100 $0.002800/sec ($10.08/hr); H100 $0.001525/sec ($5.49/hr); 2x H100 $0.003050/sec ($10.98/hr); 4x H100 $0.006100/sec ($21.96/hr); 8x H100 $0.012200/sec ($43.92/hr); H200 $0.001525/sec ($5.49/hr); 2x H200 $0.003050/sec ($10.98/hr); 4x H200 $0.006100/sec ($21.96/hr); 8x H200 $0.012200/sec ($43.92/hr); 4x A100 $0.005600/sec ($20.16/hr); 8x A100 $0.011200/sec ($40.32/hr); 4x L40S $0.003900/sec ($14.04/hr); 8x L40S $0.007800/sec ($28.08/hr). Multi-GPU tiers above 1x require committed spend contracts. Public models billed by time or per-input/output; private models on dedicated hardware billed for all time online (including idle). Enterprise & volume discounts available.

Cross-references

This record references no other records.