SnapEdit SDK

Official TypeScript / JavaScript SDK for the SnapEdit API — 40+ AI image & video editing models through a single API key: background removal, object/text/logo erasing, upscaling, restoration, image generation, virtual try-on, video enhancement, and more.

• ✅ Full TypeScript types for every endpoint
• ✅ Works in Node.js 18+, Bun, Deno, edge runtimes, and the browser
• ✅ Accepts URLs, local file paths, Blob, Buffer, and Uint8Array inputs
• ✅ Built-in retries, timeouts, and rate-limit tracking
• ✅ One-call helpers for async tasks (try-on, video, PDF) with automatic polling

Installation

npm install snapedit-api
# or: pnpm add snapedit-api / yarn add snapedit-api / bun add snapedit-api

Quick start

Get an API key from the SnapEdit dashboard (keys start with sk-snap-).

import { SnapEdit } from "snapedit-api";

const client = new SnapEdit({ apiKey: process.env.SNAPEDIT_API_KEY! });

// Remove the background from an image URL
const result = await client.remove.background({
  input_image: "https://img.snapedit.app/portrait.jpg",
});

console.log(result.data[0].url); // → https://outputs.snapedit.app/...

CommonJS works too:

const { SnapEdit } = require("snapedit-api");

Inputs: URL, path, or binary

Every input_image (and other file fields) accepts:

| Input | Example | | ----------------- | --------------------------------------------------- | | Remote URL | "https://example.com/cat.jpg" | | Local file path* | "./photos/cat.jpg" | | Blob / File | new Blob([bytes]) / a browser File | | Buffer | fs.readFileSync("./cat.jpg") | | Uint8Array | new Uint8Array(bytes) |

* Local file paths are read from disk in Node.js only.

Configuration

const client = new SnapEdit({
  apiKey: "sk-snap-...",         // required
  baseURL: "https://api.snapedit.app/v1", // optional override
  authScheme: "bearer",          // "bearer" (default) or "api-key" header
  timeout: 60_000,               // per-request timeout in ms
  maxRetries: 2,                 // retries on 429 / 5xx / network errors
  defaultHeaders: {},            // extra headers on every request
});

Examples by category

Removal

await client.remove.background({ input_image: url });            // remove bg
await client.remove.background({ input_image: url, output_type: "mask" });
await client.remove.objects({ input_image: url, erase_mode: "super" });
await client.remove.text({ input_image: url, input_mask: maskUrl });
await client.remove.logo({ input_image: url });                 // auto-detects logo
await client.remove.reflection({ input_image: url });

Detection (then erase what you found)

const det = await client.detect.objects({ input_image: url, lang: "vi" });
const removed = await client.remove.objects({
  input_image: url,
  original_session_id: det.session_id,
  mask_objects: det.detected_objects.map((o) => o.object_type).join(","),
});

Enhance / restore / colorize

await client.enhance.image({ input_image: url, zoom_factor: 4 });     // 4× upscale
await client.enhance.imagePro({ input_image: url, zoom_factor: 2 });  // higher quality
await client.enhance.restore({ input_image: oldPhotoUrl });          // fix old photos
await client.enhance.colorize({ input_image: bwPhotoUrl });          // add color

Generation

await client.generate.zimage({ prompt: "a fox in a snowy forest", aspect_ratio: "16:9" });
await client.generate.qwen({ prompt: "product photo of a coffee mug" });
await client.generate.headshot({ input_image: selfieUrl, prompt: "corporate suit, studio light" });

AI prompt editing

await client.edit.edit({
  input_image: url,
  prompt: "make it a sunset scene",
  mode: "editing",
});

await client.edit.edit({
  input_image: url,
  input_mask: maskUrl,
  prompt: "replace with a red car",
  mode: "inpaint",
});

Beauty & style

await client.beauty.retouchSkin({ input_image: portraitUrl });
await client.beauty.makeup({ input_image: portraitUrl, style: "natural_01" });
await client.beauty.hairstyle({ input_image: portraitUrl, style: "bob_02" });

Async tasks (virtual try-on, video, PDF)

Async endpoints return a task_id. Use the .run() helper to create and poll to completion in one call:

const task = await client.tryOn.run(
  {
    model_image: personUrl,
    cloth_image: shirtUrl,
    cloth_type: "upper",
    hd_mode: true,
  },
  { onProgress: (t) => console.log(t.status, t.progress) }
);

console.log(task.download_url);

Or drive the task manually:

const { task_id } = await client.tryOn.create({ /* ... */ });
const status = await client.tryOn.get(task_id);
// await client.tryOn.cancel(task_id);

Video and PDF work the same way (upload + poll handled for you):

const video = await client.video.enhance("./clip.mp4", { zoom_factor: "2K" });
const pdf = await client.files.removeLogoPdf("./document.pdf");

Error handling

All non-2xx responses throw a SnapEditError; async task failures throw a SnapEditTaskError.

import { SnapEditError } from "snapedit-api";

try {
  await client.remove.background({ input_image: url });
} catch (err) {
  if (err instanceof SnapEditError) {
    console.error(err.status, err.type, err.message);
    if (err.isRateLimit) console.log(`retry after ${err.retryAfter}s`);
  }
}

Rate limits

Default limit is 60 requests/minute per API key. After any call you can inspect the latest limits:

console.log(client.rateLimit);
// { limit: 60, remaining: 58, reset: "2026-04-28T10:00:00Z" }

The client automatically retries 429 and 5xx responses (respecting retry-after) up to maxRetries.

OpenAI SDK compatibility

The API is also compatible with the OpenAI SDK for generation/editing operations — point baseURL at SnapEdit:

import OpenAI from "openai";
const client = new OpenAI({ apiKey: "sk-snap-...", baseURL: "https://api.snapedit.app/v1" });

API coverage

| Resource | Methods | | ----------------- | ------- | | client.detect | objects, text, wires | | client.remove | objects, text, logo, wires, background, backgroundGraphic, reflection, mirror | | client.enhance | image, imagePro, art, restore, restorePro, colorize, colorizePro, lightRestore, backlitFix, nightFlash | | client.generate | zimage, qwen, art, background, headshot, sticker | | client.edit | edit, multi | | client.beauty | retouchSkin, makeup, hairstyle | | client.utility | outpaint, poseSuggest, aiDetect | | client.tryOn | create, get, cancel, wait, run | | client.video | enhance, removeLogo, getEnhanceStatus, getRemoveLogoStatus | | client.files | removeLogoPdf, getRemoveLogoPdfStatus | | client.health() | Unauthenticated health check |

Documentation

• 📖 Full API reference
• 🔑 Get an API key
• 🧪 API playground

License

MIT © SnapEdit