Replicate - Run AI with an API

replicate.com

Run open-source machine learning models with a cloud API

收录于 1 年前更新于 1 年前正常收录

详细内容

Replicate is joining Cloudflare

Run AI with an API.

Run and fine-tune models. Deploy custom models. All with one line of code.

Get started for free

NodePythonHTTP

import Replicate from

"replicate"

const replicate = new Replicate({

  auth: process.env.REPLICATE_API_TOKEN

})

const model =

"black-forest-labs/flux-2-pro"

const input = {

  prompt:

"a poolside patio at sunset with vin

};

const [output] = await replicate.run(model, { input });

console.log(output);

A poolside patio at sunset with vintage lounge chairs.

black-forest-labs/flux-2-pro

A soft armchair shaped like a peeled banana.

google/nano-banana-pro

A woman relaxing in a french bookstore.

bytedance/seedream-4

A futuristic robot looking into the distance.

black-forest-labs/flux-pro

An abstract painting of a sunrise.

black-forest-labs/flux-pro

With Replicate you can

Generate imagesGenerate speechGenerate musicRestore imagesGenerate videos from imagesLarge Language Models (LLMs)Caption Images

\
\
\
\
google/imagen-4-fast\
\
Use this fast version of Imagen 4 when speed and cost are more important than quality\
\
2.3Mruns\
\
Official \
\
\
\
bytedance/seedream-4\
\
Unified text-to-image generation and precise single-sentence editing at up to 4K resolution\
\
15Mruns\
\
Official \
\
\
\
black-forest-labs/flux-1.1-pro\
\
Faster, better FLUX Pro. Text-to-image model with excellent image quality, prompt adherence, and output diversity.\
\
64.6Mruns\
\
Official \
\
\
\
ideogram-ai/ideogram-v3-turbo\
\
Turbo is the fastest and cheapest Ideogram v3. v3 creates images with stunning realism, creative designs, and consistent styles\
\
4.8Mruns\
\
Official \
\
\
\
qwen/qwen-image\
\
An image generation foundation model in the Qwen series that achieves significant advances in complex text rendering.\
\
1.1Mruns\
\
Official \
\
\
\
black-forest-labs/flux-schnell\
\
The fastest image generation model tailored for local development and personal use\
\
557.8Mruns\
\
Official \
\
\
\
google/nano-banana\
\
Google's latest image editing model in Gemini 2.5\
\
47.3Mruns\
\
Official \
\
\
\
google/imagen-4\
\
Google's Imagen 4 flagship model\
\
6.1Mruns\
\
Official

fermatresearch

Thousands of models contributed by our community

All the latest models are on Replicate. They’re not just demos — they all actually work and have production-ready APIs.

AI shouldn’t be locked up inside academic papers and demos. Make it real by pushing it to Replicate.

Explore models

Push a model

\
\
\
\
prunaai/z-image-turbo\
\
Z-Image Turbo is a super fast text-to-image model of 6B parameters developed by Tongyi-MAI.\
\
36.4Kruns \
\
lightricks/ltx-2-retake\
\
Take any shot and edit specific sections. Rephrase, change the action, camera angles and more\
\
284runs\
\
Official \
\
\
\
google/gemini-3-pro\
\
Google's most advanced reasoning Gemini model\
\
13.1Kruns\
\
Official \
\
\
\
black-forest-labs/flux-2-pro\
\
High-quality image generation and editing with support for eight reference images\
\
101.3Kruns\
\
Official \
\
\
\
openai/gpt-5.1\
\
The best model for coding and agentic tasks with configurable reasoning effort.\
\
53.8Kruns\
\
Official \
\
\
\
google/nano-banana-pro\
\
Google's state of the art image generation and editing model 🍌🍌\
\
1.4Mruns\
\
Official

\
\
\
\
qwen/qwen-image-edit-plus-lora\
\
Qwen Image Edit 2509 LoRA explorer, uses HuggingFace URLs to load any safetensor\
\
54.4Kruns\
\
Official \
\
minimax/hailuo-2.3\
\
A high-fidelity video generation model optimized for realistic human motion, cinematic VFX, expressive characters, and strong prompt and style adherence across both text-to-video and image-to-video workflows\
\
13.1Kruns\
\
Official \
\
google/veo-3.1\
\
New and improved version of Veo 3, with higher-fidelity video, context-aware audio, reference image and last frame support\
\
119.2Kruns\
\
Official \
\
\
\
openai/sora-2\
\
OpenAI's Flagship video generation with synced audio\
\
123.1Kruns\
\
Official \
\
\
\
bytedance/seedream-4\
\
Unified text-to-image generation and precise single-sentence editing at up to 4K resolution\
\
15Mruns\
\
Official \
\
\
\
philz1337x/crystal-upscaler\
\
High-precision image upscaler optimized for portraits, faces and products. One of the upscale modes powered by Clarity AI. X:https://x.com/philz1337x\
\
210.8Kruns\
\
Official

How it works

You can get started with any model with just one line of code. But as you do more complex things, you can fine-tune models or deploy your own custom code.

Run models

Our community has already published thousands of models that are ready to use in production. You can run these with one line of code.

Explore models

import replicate

output = replicate.run(

  "black-forest-labs/flux-dev",

  input={

    "aspect_ratio": "1:1",

    "num_outputs": 1,

    "output_format": "jpg",

    "output_quality": 80,

    "prompt": "An astronaut riding a rainbow unicorn, cinematic, dramatic",

print(output)

Fine-tune models with your own data

You can improve models with your own data to create new models that are better suited to specific tasks.

Image models like SDXL can generate images of a particular person, object, or style.

Fine-tune image models

Train a model:

training = replicate.trainings.create(

  destination="mattrothenberg/drone-art"

  version="ostris/flux-dev-lora-trainer:e440909d3512c31646ee2e0c7d6f6f4923224863a6a10c494606e79fb5844497",

  input={

    "steps": 1000,

    "input_images":

https://example.com/images.zip

    "trigger_word": "TOK",

},

This will result in a new model:

mattrothenberg/drone-art

Fantastical images of drones on land and in the sky

0runs

mattrothenberg/drone-art

Fantastical images of drones on land and in the sky

0runs

Then, you can run it with one line of code:

output = replicate.run(

  "mattrothenberg/drone-art:abcde1234...",

  input={"prompt": "a photo of TOK forming a rainbow in the sky"}),

Deploy custom models

You aren’t limited to the models on Replicate: you can deploy your own custom models using Cog, our open-source tool for packaging machine learning models.

Cog takes care of generating an API server and deploying it on a big cluster in the cloud. We scale up and down to handle demand, and you only pay for the compute that you use.

Learn more

First, define the environment your model runs in with cog.yaml:

build:

  gpu: true

  system_packages:

    - "libgl1-mesa-glx"

    - "libglib2.0-0"

  python_version: "3.10"

  python_packages:

    - "torch==1.13.1"

predict: "predict.py:Predictor"

Next, define how predictions are run on your model with predict.py:

from cog import BasePredictor, Input, Path

import torch

class Predictor(BasePredictor):

  def setup(self):

      """Load the model into memory to make running multiple predictions efficient"""

      self.model = torch.load("./weights.pth")

  # The arguments and types the model takes as input

  def predict(self,

        image: Path = Input(description="Grayscale input image")

  ) -> Path:

      """Run a single prediction on the model"""

      processed_image = preprocess(image)

      output = self.model(processed_image)

      return postprocess(output)

Scale on Replicate

Thousands of businesses are building their AI products on Replicate. Your team can deploy an AI feature in a day and scale to millions of users, without having to be machine learning experts.

Learn more about our enterprise plans

Automatic scale

If you get a ton of traffic, Replicate scales up automatically to handle the demand. If you don't get any traffic, we scale down to zero and don't charge you a thing.

CPU$0.000100/sec
Nvidia T4 GPU$0.000225/sec
Nvidia L40S GPU$0.000975/sec
2x Nvidia L40S GPU$0.001950/sec
Nvidia A100 (80GB) GPU$0.001400/sec
8x Nvidia A100 (80GB) GPU$0.011200/sec
Learn more about pricing

Pay for what you use

Replicate only bills you for how long your code is running. You don't pay for expensive GPUs when you're not using them.

Forget about infrastructure

Deploying machine learning models at scale is hard. If you've tried, you know. API servers, weird dependencies, enormous model weights, CUDA, GPUs, batching.

06:58 GMT07:15 GMT07:32 GMT07:49 GMT08:06 GMT08:23 GMT08:40 GMT08:58 GMT014274154

Prediction throughput(requests per second)

Logging & monitoring

Metrics let you keep an eye on how your models are performing, and logs let you zoom in on particular predictions to debug how your model is behaving.

Replicate logo

Imagine what you can build

Autonomous RobotsZero-shot autonomous robots with open source models Paint with AIAn iPad app that lets you paint with AI emojis.shAI Emojis Never Gonna Give You UpVideo fine tunes on Replicate Language Model CLILanguage model command line interface

Imagine

Autonomous RobotsZero-shot autonomous robots with open source models

what

you

Paint with AIAn iPad app that lets you paint with AI

can

emojis.shAI Emojis

Never Gonna Give You UpVideo fine tunes on Replicate

build.

Language Model CLILanguage model command line interface

With Replicate and tools like Next.js and Vercel, you can wake up with an idea and watch it hit the front page of Hacker News by the time you go to bed.

Get started

Replicate glyphDownload logo glyph as SVG

Download wordmark as SVG

Generate your own goo

Replicate glyphDownload logo glyph as SVG

Download wordmark as SVG

Generate your own goo

网站信息

官网地址: https://replicate.com/
域名: replicate.com
收录路径: replicate