DaDaScribe advanced speech-to-text transcription & translation

DaDaScribe API Skill

Use the official DaDaScribe API (https://api.dadascribe.com/v1) for high-quality AI transcription.

Base URL: https://api.dadascribe.com/v1
API version: 1.0.1
Official docs: https://api.dadascribe.com/docs
OpenAPI: https://api.dadascribe.com/openapi.json
Get API key: https://www.dadascribe.com/account/api.php (keys start with dds_)

The official Python wrapper lives in this same repository (dadascribe package). Prefer the HTTP API below when the package is not installed; it works with any language that can make HTTP requests.

Authentication

Every authenticated request requires:

Authorization: Bearer dds_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Never hard-code the key. Load it from the environment variable DADASCRIBE_API_KEY (or ask the user for it). Output file download URLs do not require authentication.

Core Workflow

1. Submit one or more sources → receive a batch id.
2. Poll /status until status is "complete" (or "error").
3. Download the .txt / .srt files from the urls array within 1 hour (they expire for privacy).

1. Submit transcription – POST /transcribe

JSON body (preferred for URLs):

{
  "source": "https://www.youtube.com/watch?v=VIDEO_ID",
  "source-language": "en",
  "destination-language": "es,it,fr",
  "diarization": "Host,Guest"
}

• source (required for JSON unless s3_token is used): single string or array of up to 10 items.
  Accepted values: full YouTube URL, Shorts, youtu.be, 11-char video ID, or a direct HTTPS audio/video URL.
• s3_token (optional): token from /v1/upload/complete instead of source / file.
• Direct file URLs and uploads allow audio and video (max 10 GB; PHP multipart file is still capped near 480M — use /v1/upload/* for larger files).
• source-language (required): ISO-style code (en, es, fr, zh, ja, …). 100+ supported.
• destination-language (optional): comma-separated list, max 5, must not include the source language.
• diarization (optional): comma-separated speaker names, minimum 2.

Multipart (for a single local file upload):

file=@/path/to/audio.mp3
data={"source-language":"en","destination-language":"es"};type=application/json

or individual form fields.

Success response:

{
  "status": "ok",
  "id": "a1B2c3D4e5F6g7H8",
  "count": 1
}

Limits:

• Max 10 sources per request
• Max 10 hours per source
• Duplicates are automatically removed
• Rate limit: 60 requests/minute per API key

2. Poll status – POST /status

{
  "id": "a1B2c3D4e5F6g7H8"
}

Possible responses:

Rate limit: 120 requests/minute. Poll every 5–15 seconds.

3. Download results – GET /output/{id}/{filename}

No auth header needed. Example URLs returned in a complete response:

https://api.dadascribe.com/v1/output/a1B2c3D4e5F6g7H8/xxxx.txt
https://api.dadascribe.com/v1/output/a1B2c3D4e5F6g7H8/xxxx.srt
https://api.dadascribe.com/v1/output/a1B2c3D4e5F6g7H8/xxxx_es.srt

• .txt = plain transcript
• .srt = timed subtitles
• _{lang}.srt = translated subtitles

Files disappear after 1 hour. Rate limit: 600/min per IP.

Python reference implementation (requests)

import os
import time
import requests

API_KEY = os.environ["DADASCRIBE_API_KEY"]
BASE = "https://api.dadascribe.com/v1"
HEADERS = {
    "Authorization": f"Bearer {API_KEY}",
    "Content-Type": "application/json",
}

def transcribe(source, source_language="en", destination_language=None, diarization=None):
    payload = {"source": source, "source-language": source_language}
    if destination_language:
        payload["destination-language"] = destination_language
    if diarization:
        payload["diarization"] = diarization

    r = requests.post(f"{BASE}/transcribe", json=payload, headers=HEADERS, timeout=30)
    r.raise_for_status()
    data = r.json()
    if data.get("status") != "ok":
        raise RuntimeError(data)
    return data["id"]

def wait_until_complete(job_id, poll_interval=8):
    while True:
        r = requests.post(f"{BASE}/status", json={"id": job_id}, headers=HEADERS, timeout=15)
        r.raise_for_status()
        data = r.json()
        status = data.get("status")
        if status == "complete":
            return data["urls"]
        if status == "error":
            raise RuntimeError(data.get("message", "unknown error"))
        time.sleep(poll_interval)

def download(urls, out_dir="."):
    paths = []
    for url in urls:
        name = url.rsplit("/", 1)[-1]
        path = os.path.join(out_dir, name)
        with requests.get(url, stream=True, timeout=60) as r:
            r.raise_for_status()
            with open(path, "wb") as f:
                for chunk in r.iter_content(8192):
                    f.write(chunk)
        paths.append(path)
    return paths

# Example usage
job_id = transcribe(
    source="https://www.youtube.com/watch?v=dQw4w9wgccc",
    source_language="en",
    destination_language="es,it",
    diarization="Host,Guest",
)
urls = wait_until_complete(job_id)
files = download(urls)
print("Downloaded:", files)

Using the official Python package (optional)

pip install -e .          # from this repo
# or
pip install git+https://github.com/PatzEdi/dadascribe-api-python.git

from dadascribe import ScribeAPIWrapper
import os

w = ScribeAPIWrapper(os.environ["DADASCRIBE_API_KEY"])
result = w.transcribe(
    source="https://www.youtube.com/watch?v=...",
    source_language="en",
    destination_language="es,it",
)

CLI (after install):

dadascribe --source "https://youtube.com/watch?v=..."
dadascribe --status 
dadascribe --download 

Error handling

Always surface the message field from error JSON to the user.

Best practices for agents

• Confirm the user has a valid API key and account balance before starting large jobs.
• Prefer YouTube / direct URLs over uploading large local files when possible.
• Respect rate limits; never spin in a tight poll loop.
• Download results immediately after complete.
• When returning results to the user, prefer the plain .txt for readability and the .srt for subtitles.
• For multi-language requests, list the generated _{lang}.srt files clearly.
• Never store or log the full API key.

This skill gives any agent complete, production-ready access to the DaDaScribe transcription service.