API Intermediate

The D-ID Talks API lets you programmatically create talking head videos from source images and text or audio. This lesson covers authentication, the talk creation workflow, status polling, and building robust integrations.

Authentication

D-ID uses API key authentication via Basic Auth. Your API key serves as the username with an empty password, Base64-encoded in the Authorization header.

import requests

API_KEY = "your_api_key"
BASE_URL = "https://api.d-id.com"
headers = {
    "Authorization": f"Basic {API_KEY}",
    "Content-Type": "application/json"
}

# Create a talk
payload = {
    "source_url": "https://example.com/face.jpg",
    "script": {
        "type": "text",
        "input": "Hello from the D-ID API!",
        "provider": {"type": "microsoft", "voice_id": "en-US-JennyNeural"}
    }
}
resp = requests.post(f"{BASE_URL}/talks", json=payload, headers=headers)
talk_id = resp.json()["id"]

Talk Creation Workflow

1. POST /talks

   Create a new talk with source image URL and script. Returns a talk ID.

2. GET /talks/{id}

   Poll for status. States: created → started → done (or error).

3. Download result

   When status is "done", the response includes result_url with the MP4 video.

Input Options