API Documentation | XTCStudio

Quick Start

Up and running in 60 seconds

Install the SDK, authenticate, and make your first API call.

1. Install

pip install xtcstudio

# No installation needed. Use curl directly.

2. Authenticate

from xtcstudio import XTC

client = XTC(api_key="your-api-key")

# Pass your key via the Authorization header
curl https://api.xtcstudio.com/v1/jobs \
  -H "Authorization: Bearer xtc_your_key_here"

3. First call

job = client.transcribe(
    url="https://youtube.com/watch?v=example",
    engine="studio_pro",
    diarize=True
)
print(job.status)  # "processing"

curl -X POST https://api.xtcstudio.com/v1/transcribe \
  -H "Authorization: Bearer xtc_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{
    "url": "https://youtube.com/watch?v=example",
    "engine": "studio_pro",
    "diarize": true
  }'

Reference

Endpoints

Five endpoints cover the full pipeline: transcribe, extract, organize, check status, and export.

Method	Endpoint	Description
POST	`/v1/transcribe`	Submit a URL or file for transcription
POST	`/v1/extract`	Run AI extraction on a transcript
POST	`/v1/organize`	Cluster extractions by semantic similarity
GET	`/v1/jobs/{id}`	Check job status
GET	`/v1/export/{id}`	Download results in specified format

Authentication

API Keys

All requests require an API key passed in the Authorization header.

Header format

Authorization: Bearer xtc_your_key_here

Getting your key

1. Sign up at xtcstudio.com.
2. Open the Dashboard and navigate to Settings > API Keys.
3. Click Generate Key. Copy it immediately. Keys are only shown once.

Keep your key secret

Never expose API keys in client-side code, public repositories, or logs. Use environment variables or a secrets manager.

Full Example

End-to-end pipeline

Transcribe a video, extract quotes, and export the results as JSON.

from xtcstudio import XTC
import time

client = XTC(api_key="xtc_your_key")

# Transcribe
job = client.transcribe(
    url="https://youtube.com/watch?v=example",
    engine="studio_pro",
    diarize=True
)

# Wait for completion
result = client.wait(job.id)

# Extract quotes
extraction = client.extract(
    transcript_id=result.id,
    mode="quotes",
    llm="claude",
    model="sonnet"
)

# Export
output = client.export(extraction.id, format="json")
print(output)

import XTC from 'xtcstudio';

const client = new XTC({ apiKey: 'xtc_your_key' });

async function run() {
  // Transcribe
  const job = await client.transcribe({
    url: 'https://youtube.com/watch?v=example',
    engine: 'studio_pro',
    diarize: true
  });

  // Wait for completion
  const result = await client.wait(job.id);

  // Extract quotes
  const extraction = await client.extract({
    transcriptId: result.id,
    mode: 'quotes',
    llm: 'claude',
    model: 'sonnet'
  });

  // Export
  const output = await client.export(extraction.id, { format: 'json' });
  console.log(output);
}

run();

# 1. Submit transcription
curl -X POST https://api.xtcstudio.com/v1/transcribe \
  -H "Authorization: Bearer xtc_your_key" \
  -H "Content-Type: application/json" \
  -d '{"url":"https://youtube.com/watch?v=example","engine":"studio_pro","diarize":true}'

# Response: {"id":"job_abc123","status":"processing"}

# 2. Poll for completion
curl https://api.xtcstudio.com/v1/jobs/job_abc123 \
  -H "Authorization: Bearer xtc_your_key"

# Response: {"id":"job_abc123","status":"completed","transcript_id":"tx_abc123"}

# 3. Extract quotes
curl -X POST https://api.xtcstudio.com/v1/extract \
  -H "Authorization: Bearer xtc_your_key" \
  -H "Content-Type: application/json" \
  -d '{"transcript_id":"tx_abc123","mode":"quotes","llm":"claude","model":"sonnet"}'

# Response: {"id":"ext_abc123","status":"completed"}

# 4. Export results
curl https://api.xtcstudio.com/v1/export/ext_abc123?format=json \
  -H "Authorization: Bearer xtc_your_key"

Extraction

Extraction Modes

Specify the mode parameter in your extract call to control what the AI pulls from the transcript.

Quotes

mode: "quotes"

Extracts notable quotes with speaker attribution, timestamps, and confidence scores.

Wisdom

mode: "wisdom"

Pulls key insights, lessons, and actionable takeaways from the content.

Dialogues

mode: "dialogues"

Identifies and structures meaningful exchanges between speakers.

Book English

mode: "book_english"

Rewrites spoken content into polished, publication-ready prose.

Custom

mode: "custom"

Pass your own prompt. Add prompt: "..." to define exactly what to extract.

Custom mode exampleextraction = client.extract(
    transcript_id=result.id,
    mode="custom",
    prompt="Extract all mentions of dates, deadlines, and time-sensitive commitments.",
    llm="claude",
    model="sonnet"
)

Webhooks

Async Callbacks

Instead of polling for job status, register a webhook URL to receive a POST request when a job completes. Set the webhook_url parameter on any transcribe or extract call.

Payload example

{
  "event": "job.completed",
  "job_id": "job_abc123",
  "type": "transcription",
  "status": "completed",
  "result_url": "/v1/export/job_abc123"
}

Verification

Every webhook request includes an X-XTC-Signature header. Verify it against your webhook secret to confirm authenticity.

Tier	Requests / min	Concurrent jobs
Free	10	2
Pro	60	10
Business	200	50

Responses

Response Format

All responses are JSON. Here is an example extraction response.

{
  "id": "ext_abc123",
  "status": "completed",
  "mode": "quotes",
  "items": [
    {
      "text": "The only way to do great work is to love what you do.",
      "speaker": "Speaker 1",
      "timestamp": "00:14:23",
      "confidence": 0.94
    }
  ],
  "metadata": {
    "source_url": "https://youtube.com/watch?v=example",
    "duration": 3847,
    "word_count": 12450
  }
}

Build on XTC.