Stop building scrapers.Start shipping.

Word-level transcripts from any video URL, returned in under a second. Works with every major platform.

Works with

YouTube
TikTok
Instagram
Facebook
X
YouTube
TikTok
Instagram
Facebook
X
YouTube
TikTok
Instagram
Facebook
X

Everything you need, nothing you don't

Built for developers who ship fast and need transcription to just work.

Sub-Second Extraction

Get a full transcript back in under a second. No queues, no waiting, no cold starts.

Every Major Platform

YouTube, TikTok, Instagram, Facebook, X, and Vimeo — all from a single endpoint.

Word-Level Timestamps

Every word comes with its own start and end time — perfect for subtitles, search, and clipping.

100+ Languages

We use native captions when available, and fall back to a speech model when they're not.

Private by Default

Zero data retention on every plan. We never train on your content.

Any Output Format

Plain text, SRT, VTT, JSON, or Markdown. Pick the format that fits your stack.

Vector Databases, Explained
0:00 / 1:34
YouTube · 12:04 · English
Productividad para devs
0:00 / 1:34
TikTok · 00:58 · Spanish
Universal API

Any video. Any platform. One call.

Drop a video URL. Get word-level captions back in JSON, SRT, VTT, or plain text. Same call works across YouTube, TikTok, Instagram, Facebook, X, and Vimeo — no per-platform integrations to maintain.

  • One endpoint
    Stop juggling scrapers and SDKs per platform. One HTTP call covers every major host.
  • Word-level timestamps
    Sub-second start/end on every word. Perfect for editing, search, and highlight clips.
  • Native captions first
    We use the platform's own captions when they exist (free, fast) and fall back to ASR when they don't.

Ship in minutes, not days

One request. Any platform. Your transcript is ready before your coffee gets cold.

transcribe.py
1import requests
2 
3response = requests.get(
4 "https://scriptbase.app/api/v1/transcribe",
5 params={"url": "https://youtube.com/watch?v=dQw4w9WgXcQ"},
6 headers={"X-API-Key": "YOUR_API_KEY"},
7)
8 
9print(response.json()["data"]["full_text"])
~/my-project

Teams love building
with ScriptBase

Here's what our customers say about us

We ship transcripts to Notion before our podcast episodes finish uploading. ScriptBase is the fastest thing in our stack.

JP
Jordan Patel
Founder at Podline

We publish in 40+ languages. ScriptBase was the only API that didn't fall over on our Indonesian and Vietnamese content.

SA
Sofia Alvarez
CTO at Loopwave

Dropped the Python SDK into our Airflow DAG and had transcripts flowing into Snowflake by the end of the day.

EB
Ethan Brooks
Staff Engineer at Northstar

Word-level timestamps are what sold it for us. Our highlight generator wouldn't exist without them.

PD
Priya Desai
Data Eng at Quanta

Our team cites interview moments directly from the timestamped transcripts. It's changed how we do qualitative research.

ML
Marcus Lee
Researcher at Vectorlab

I batch-drop a week of YouTube uploads into ScriptBase and walk away with captions, show notes, and newsletters done.

NK
Nora Kim
Creator & Educator

Questions? We got answers

Check out the answers to most frequently asked questions.

YouTube, TikTok, Instagram Reels, Facebook Watch, X, Vimeo, and most public video hosts. If you can share a link, we can probably transcribe it.

We use native captions when they exist and a state-of-the-art speech model otherwise. Expect 95%+ word accuracy on clean audio and strong performance on noisy real-world recordings.

No. Zero-retention mode is available on every plan and is the default on Pro and above. We never train models on your content.

JSON, plain text, SRT, VTT, and Markdown. You can request multiple formats in a single call and get word-level timestamps alongside each one.

Yes — every account includes 25 free credits. No credit card required to start.

Enterprise customers can deploy ScriptBase inside their own VPC. Reach out to sales for deployment options and pricing.

Ship transcripts today.

25 free credits. No credit card. Build a prototype before lunch.