Files

The files endpoint stores a file once so you can reference it by id in any number of chat requests instead of re-encoding the bytes into every message. The upload takes the same multipart form as the OpenAI files endpoint, so client.files.create sends it as is. The file object that comes back carries OpenDunes' own fields (filename, mime_type, size_bytes, created_at, downloadable), listed in the Files guide; id is the one to keep. Uploads, storage and downloads are free.

Upload a file#

import fs from "node:fs";
import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://opendunes.com/v1",
  apiKey: process.env.OPENDUNES_API_KEY,
});

const file = await client.files.create({
  file: fs.createReadStream("report.pdf"),
  purpose: "user_data",
});

console.log(file.id); // reference this id in chat requests

Reference it in a chat request#

Send a file content part carrying the file_id — the server inlines the bytes for you:

const completion = await client.chat.completions.create({
  model: "google/gemini-2.5-flash",
  messages: [
    {
      role: "user",
      content: [
        { type: "text", text: "Summarize this document." },
        { type: "file", file: { file_id: file.id } },
      ],
    },
  ],
});

Notes#

• Max 100 MB per file, no total per account, and no expiry: a file stays until you delete it.
• Supported types: images (png, jpeg, webp, gif), audio (mp3, wav, flac, m4a, ogg, aac), PDF, office documents (docx, xlsx, pptx), and plain text.
• The model must support the file's modality — vision for images, audio input for audio, file/PDF input for documents.
• List, retrieve, and delete are also available — see the Files guide for the full surface.

Also see#

• Files guide — endpoints, limits, and reference semantics
• Files reference