Registry / Tools / Replicate

Replicate facts

tool_replicate
Platform

Replicate

Cloud API platform for running, fine-tuning, and deploying AI models.

last verified: 2026-05-10pricing checked: 2026-05-10
01

Fact summary

Name
Replicate
Company / maker
Replicate
Entity type
Platform
Availability
Public
Pricing model
Usage Based
Pricing unit
month
Primary pricing route
Model Api
Delivery modes
APISDKWeb AppCLI
Supported platforms
Web
Data state
source backed
02

Official links

Formal source rows with URL, source type, and freshness.
LinkURLsource typechecked
Official websitereplicate.com official-site2026-05-10
Pricing pagereplicate.com/pricing official-pricing-page2026-05-10
Documentationreplicate.com/docs official-documentation2026-05-10
Changelogreplicate.com/changelog official-changelog2026-05-10
API docsreplicate.com/api official-documentation2026-05-10
03

Access and delivery

Supported platforms and delivery modes are intentionally separate.

Supported platforms

Web

Delivery modes

APISDKWeb AppCLI
04

Capabilities

Capability tags

Model ApisModel HostingImage GenerationVideo Generation
Input types
TextImageAudioVideoStructured DataFunction Call
Output types
TextImageAudioVideoStructured JsonEmbeddings
06

Pricing summary

Pricing summary

Derived from official pricing routes where available
pricingModel
Usage Based
pricingUnit
month
startingPrice
$0.00 / month
currency
USD
pricingLastChecked
2026-05-10
07

Pricing routes

Each route is a first-class record with source URL and checked date.
Enterprise and volume discounts
routeType: enterprise-sales

Official pricing says Enterprise and volume discounts are available for dedicated account management, priority support, higher GPU limits, performance SLAs, onboarding, custom models, and optimizations.

Private model hardware
routeType: api-usage

Official pricing says most private models run on dedicated hardware billed for online time and lists hardware prices per second and per hour.

Public model usage
routeType: model-api
primary route

Official pricing says public models are billed by hardware time or by input/output, with model pages showing cost estimates.

08

Pricing plans

plantrackmonthlyusage limitfeaturessourcechecked
Private models - CPU (Small)model-api0.000025per second; $0.09 per hour; CPU 1x; RAM 2GBPrivate models are billed for instance online time, including setup, idle time, and active processing2026-05-10
Private models - CPUmodel-api0.000100per second; $0.36 per hour; CPU 4x; RAM 8GBPrivate models are billed for instance online time, including setup, idle time, and active processing2026-05-10
Private models - Nvidia T4 GPUmodel-api0.000225per second; $0.81 per hour; GPU 1x; CPU 4x; GPU RAM 16GB; RAM 16GBPrivate models are billed for instance online time, including setup, idle time, and active processing2026-05-10
Private models - Nvidia L40S GPUmodel-api0.000975per second; $3.51 per hour; GPU 1x; CPU 10x; GPU RAM 48GB; RAM 65GBPrivate models are billed for instance online time, including setup, idle time, and active processing2026-05-10
Private models - Nvidia A100 (80GB) GPUmodel-api0.001400per second; $5.04 per hour; GPU 1x; CPU 10x; GPU RAM 80GB; RAM 144GBPrivate models are billed for instance online time, including setup, idle time, and active processing2026-05-10
Private models - Nvidia H100 GPUmodel-api0.001525per second; $5.49 per hour; GPU 1x; CPU 13x; GPU RAM 80GB; RAM 72GBPrivate models are billed for instance online time, including setup, idle time, and active processing2026-05-10
Private models - 2x Nvidia L40S GPUmodel-api0.001950per second; $7.02 per hour; GPU 2x; CPU 20x; GPU RAM 96GB; RAM 144GBPrivate models are billed for instance online time, including setup, idle time, and active processing2026-05-10
Private models - 2x Nvidia A100 (80GB) GPUmodel-api0.002800per second; $10.08 per hour; GPU 2x; CPU 20x; GPU RAM 160GB; RAM 288GBPrivate models are billed for instance online time, including setup, idle time, and active processing2026-05-10
Private models - 2x Nvidia H100 GPUmodel-api0.003050per second; $10.98 per hourAdditional multi-GPU H100 capacity is available with committed spend contracts2026-05-10
Private models - 4x Nvidia L40S GPUmodel-api0.003900per second; $14.04 per hourAdditional multi-GPU L40S capacity is available with committed spend contracts2026-05-10
Private models - 4x Nvidia A100 (80GB) GPUmodel-api0.005600per second; $20.16 per hourAdditional multi-GPU A100 capacity is available with committed spend contracts2026-05-10
Private models - 4x Nvidia H100 GPUmodel-api0.006100per second; $21.96 per hourAdditional multi-GPU H100 capacity is available with committed spend contracts2026-05-10
Private models - 8x Nvidia L40S GPUmodel-api0.007800per second; $28.08 per hourAdditional multi-GPU L40S capacity is available with committed spend contracts2026-05-10
Private models - 8x Nvidia A100 (80GB) GPUmodel-api0.011200per second; $40.32 per hourAdditional multi-GPU A100 capacity is available with committed spend contracts2026-05-10
Private models - 8x Nvidia H100 GPUmodel-api0.012200per second; $43.92 per hourAdditional multi-GPU H100 capacity is available with committed spend contracts2026-05-10
Nvidia T4 GPUapi-$0.000225 per second; $0.81 per hour1x GPU, 4x CPU, 16GB GPU RAM, 16GB RAM2026-05-10
Nvidia L40S GPUapi-$0.000975 per second; $3.51 per hour1x GPU, 10x CPU, 48GB GPU RAM, 65GB RAM2026-05-10
Nvidia H100 GPUapi-$0.001525 per second; $5.49 per hour1x GPU, 13x CPU, 80GB GPU RAM, 72GB RAM2026-05-10
Nvidia A100 80GB GPUapi-$0.001400 per second; $5.04 per hour1x GPU, 10x CPU, 80GB GPU RAM, 144GB RAM2026-05-10
CPUapi-$0.000100 per second; $0.36 per hour4x CPU, 8GB RAM2026-05-10
CPU Smallapi-$0.000025 per second; $0.09 per hour1x CPU, 2GB RAM2026-05-10
Public modelsapi-Pay only for what you use; most models billed by run time, with price per second varying by hardware; some models billed by input and outputThousands of open-source community models, Proprietary models, Model pages include cost estimates2026-05-10
09

Evidence and sources

source URLsource typeused forchecked
replicate.com/docs/guides/deploy-a-custom-modelofficial-documentationmodel-hosting, deployments, capability-claims2026-05-10
replicate.com/apiofficial-documentationapi-access, capability-claims2026-05-10
replicate.com/docsofficial-documentationdocumentation, capabilities2026-05-10
replicate.com/pricingofficial-pricing-pagepricing, pricing-routes2026-05-10
replicate.comofficial-siteidentity, capabilities, access2026-05-10
10

Freshness and recent changes

Facts verified
2026-05-10
Identity and core facts
Pricing checked
2026-05-10
Pricing routes and plans
Record updated
2026-05-10
Public record freshness
11

Related pages

One-hop registry pages derived from this record's modeled facts. This section does not infer peer tools.