Skip to content
API pricing

Replicate API pricing, checked at the source

Prices for Replicate's official models, read off each model's own page on Replicate with a link on every row. Estimate a monthly bill below.

10
Models priced
$0.18
Lowest input / 1M
$0.72
Lowest output / 1M
2026-10-25
Prices checked
Prices

Replicate API prices per 1M tokens

Standard prices in USD. Each row links to Replicate's pricing page and shows the day we checked it.

Replicate API prices in USD per 1M tokens, checked 2026-10-25
ModelInputCached inputOutputNotesSource
anthropic/claude-fable-5$10.00none listed$50.00Sourcechecked 2026-10-25
anthropic/claude-opus-5.5$4.00none listed$20.00Sourcechecked 2026-10-25
anthropic/claude-sonnet-5$2.00none listed$10.00Sourcechecked 2026-10-25
openai/gpt-5.6-sol$2.00none listed$10.00Sourcechecked 2026-10-25
openai/gpt-5.6-luna$1.00none listed$6.00Sourcechecked 2026-10-25
google/gemini-3.1-pro$2.00none listed$12.00Sourcechecked 2026-10-25
deepseek-ai/deepseek-r1$3.75none listed$10.00Sourcechecked 2026-10-25
moonshotai/kimi-k2.5$0.72none listed$3.60Sourcechecked 2026-10-25
deepseek-ai/deepseek-v3.1$0.672none listed$2.016Sourcechecked 2026-10-25
openai/gpt-oss-120b$0.18none listed$0.72Sourcechecked 2026-10-25

Terms on this page: Token · Prompt caching · Context window · Batch API

Estimator

Estimate your Replicate bill

Enter your monthly tokens and how they split between new input, cached input and output. The result updates as you type.

Write 40k, 250M or 5B. Input, cached input and output tokens together.

Token split (%)

100M tokens a month costs from $20.70 (openai/gpt-oss-120b) to $1,200 (anthropic/claude-fable-5).

ModelPer monthPer 1M tokens at this split
openai/gpt-oss-120b$20.70$0.207
deepseek-ai/deepseek-v3.1$73.92$0.7392
moonshotai/kimi-k2.5$86.40$0.864
openai/gpt-5.6-luna$125.00$1.25
anthropic/claude-sonnet-5$240.00$2.40
openai/gpt-5.6-sol$240.00$2.40
google/gemini-3.1-pro$250.00$2.50
deepseek-ai/deepseek-r1$406.25$4.0625
anthropic/claude-opus-5.5$480.00$4.80
anthropic/claude-fable-5$1,200$12.00

Standard prices in USD on 2026-10-25, before tax. Long-prompt tiers are left out; see the price table.

Image generation

Replicate image generation prices

Prices per image where the vendor lists one. Rows billed per megapixel or per token show how they bill instead.

Replicate image generation prices in USD, checked 2026-10-25
ModelPrice as listedPer imageNotesSource
black-forest-labs/flux-schnell$0.003 per image$0.003Listed as $3.00 per thousand images.Sourcechecked 2026-10-25
google/imagen-4-fast$0.02 per image$0.02Sourcechecked 2026-10-25
black-forest-labs/flux-3-image$0.024 per image$0.024At 1k resolution; $0.05 at 2k, $0.3035 at 4k.Sourcechecked 2026-10-25
alibaba/qwen-image-3$0.03 per image$0.03Sourcechecked 2026-10-25
black-forest-labs/flux-1.1-pro$0.04 per image$0.04Sourcechecked 2026-10-25
bytedance/seedream-4.5$0.04 per image$0.04Sourcechecked 2026-10-25
google/imagen-4$0.04 per image$0.04Sourcechecked 2026-10-25
google/imagen-4-ultra$0.06 per image$0.06Sourcechecked 2026-10-25
ideogram-ai/ideogram-v3-quality$0.09 per image$0.09Sourcechecked 2026-10-25
google/nano-banana-pro$0.15 per image$0.15Same at 2K; $0.30 at 4K.Sourcechecked 2026-10-25
black-forest-labs/flux-2-dev$0.014 per megapixelbilled per megapixelgo_fast variant is $0.012 per megapixel.Sourcechecked 2026-10-25
black-forest-labs/flux-2-pro$0.015 per megapixelbilled per megapixelPlus $0.015 per run and $0.015 per input image megapixel.Sourcechecked 2026-10-25

Write 500, 10k or 1M.

1k images a month costs from $3.00 (black-forest-labs/flux-schnell) to $150.00 (google/nano-banana-pro).

ModelPer monthPer image
black-forest-labs/flux-schnell$3.00$0.003
google/imagen-4-fast$20.00$0.02
black-forest-labs/flux-3-image$24.00$0.024
alibaba/qwen-image-3$30.00$0.03
black-forest-labs/flux-1.1-pro$40.00$0.04
bytedance/seedream-4.5$40.00$0.04
google/imagen-4$40.00$0.04
google/imagen-4-ultra$60.00$0.06
ideogram-ai/ideogram-v3-quality$90.00$0.09
google/nano-banana-pro$150.00$0.15

List prices in USD on 2026-10-25, before tax, free tiers and add-ons. Rows billed another way are in the price table only.

Billing

How the bill is worked out

Two ways to bill

Most public models are billed by the seconds they run, at a per-second rate set by the hardware. Some official models are billed by input and output, such as tokens or images.

Private models pay for idle time

Private models run on dedicated hardware, so you pay for setup, idle and active time. Fast booting fine-tunes only bill active time.

Hardware rates

Examples: Nvidia T4 $0.000225 per second, L40S $0.000975 per second, H100 $0.001525 per second ($5.49 per hour).
FAQ

Replicate API pricing questions

How much does the Replicate API cost?

Replicate bills its language models per million tokens, with separate prices for input, cached input and output. On 2026-10-25, standard prices ran from $0.18 (openai/gpt-oss-120b) to $10.00 (anthropic/claude-fable-5) per 1M input tokens, and $0.72 (openai/gpt-oss-120b) to $50.00 (anthropic/claude-fable-5) per 1M output tokens. Image generation runs from $0.003 (black-forest-labs/flux-schnell) to $0.15 (google/nano-banana-pro) per image.

Does Replicate offer a batch discount?

Its pricing page lists no batch price for the rows here. Most public models are billed by the seconds they run, at a per-second rate set by the hardware. Some official models are billed by input and output, such as tokens or images.

Where do these prices come from?

From Replicate's own pricing page, https://replicate.com/pricing, checked on 2026-10-25. Every row links to it. When a price cannot be confirmed, the row says to check the vendor's page instead of showing a number.

How does Replicate bill?

Two ways. Most public models are billed by the seconds they run, at a per-second rate set by the hardware, for example $0.001525 a second on an Nvidia H100. Official models, like the ones on this page, are billed by input and output, such as tokens or images.

Related

Compare and plan

Building with the Replicate API? OceanLab builds AI into business workflows, with people kept in charge. AI automation →

START YOUR CONVERSATION

Your Next Mission-Critical
System Starts Here

Speak with a senior engineer about your requirements. No sales process, just a direct technical conversation about what you need to build.

OceanLab

Enterprise Software Engineering - Human + AI

Languages

Locations

Global Headquarters

21 Merchant Road, 3rd Floor, Singapore 058267

Main corporate office and leadership operations

North America Office

548 Market Street, PMB 12472, San Francisco, CA 94104

Client operations and business development

India Office

402 Shivalik Shilp II, SG Highway, Ahmedabad 380015

Engineering and delivery center

© 2026 - Copyright. All Rights reserved OceanLab