Server upload route

A route that takes a file from the browser, converts it with your key, and returns signed output URLs.

This route sits between your web app and convt so the API key stays on the server. It uses the standard Request and Response objects, so the same handler runs on Bun, Deno, Node 20+ frameworks, Cloudflare Workers and Next.js route handlers. It imports convert and formatFor from the Node helper.

import { convert, formatFor } from "./convt.mjs";

const MAX_BYTES = 50 * 1024 * 1024; // your limit, below convt's 2 GB
const TARGETS = new Set(["webp", "png", "jpeg", "pdf"]);

export async function handleConvert(request) {
  // Authenticate your own user here before spending your convt budget.

  const form = await request.formData();
  const file = form.get("file");
  const to = String(form.get("to") ?? "");
  if (!(file instanceof File)) return new Response("file is required", { status: 400 });
  if (!TARGETS.has(to)) return new Response("unsupported target", { status: 400 });
  if (file.size === 0 || file.size > MAX_BYTES)
    return new Response("file too large", { status: 413 });

  try {
    const outputs = await convert({
      bytes: new Uint8Array(await file.arrayBuffer()),
      from: await formatFor(file.name),
      to,
    });
    return Response.json({ outputs });
  } catch (error) {
    console.error(error);
    return new Response(error.message, { status: 502 });
  }
}

// Bun: bun server.mjs
if (typeof Bun !== "undefined") {
  Bun.serve({
    port: 3000,
    fetch: (request) =>
      new URL(request.url).pathname === "/api/convert" && request.method === "POST"
        ? handleConvert(request)
        : new Response("not found", { status: 404 }),
  });
}

The browser side is the form from the Quick start:

const body = new FormData();
body.append("file", fileInput.files[0]);
body.append("to", "webp");
const res = await fetch("/api/convert", { method: "POST", body });
const { outputs } = await res.json();
link.href = outputs[0].url;

The signed URLs expire after 5 minutes. To let users download later, store the job id and call GET /v1/jobs/{id}/download for fresh URLs when they click, any time in the 24 hours before the job expires.

Browser apps explains why the upload goes through your server.