Home / Topics / Predictions

Input files

Some models accept files as input, like images, audio, or video, zip files, PDFs, etc.

There are multiple ways to use files as input when running a model on Replicate. You can provide a file as input using a URL, a local file on your computer, or a base64-encoded object.

Option 1: Hosted file

Use a URL to provide a hosted file:

const fileInput = "https://example.com/path/to/your/file";

This is useful if you already have a file hosted somewhere on the internet.

Option 2: Local file

You can provide Replicate with a Blob, File, or Buffer object, and the library will handle the upload for you. This will work for files up to 100MB:

import { readFile } from "node:fs/promises";
const fileInput = await readFile("./path/to/your/file");

Option 3: Data URI

Create a data URI consisting of the base64 encoded data for your file. This is only recommended if the file is less than 1MB:

import { readFile } from "node:fs/promises";
const data = (await readFile("./path/to/your/file")).toString("base64");
const fileInput = `data:application/octet-stream;base64,${data}`;

Using the file input

Once you have your file input ready, you can use it in your prediction:

const input = { file: fileInput };
const output = await replicate.run("your-model-id", { input });
Next: Output files