JP Transcribe
Free · CPU · REST API

Japanese speech → text, in one API call

Transcribe Japanese audio and optionally translate to English. Built for stream overlays, mobile apps, and backend automation.

Community-powered: run a worker on your CPU for free API access — our server coordinates, you transcribe.

Get free API key Read the docs
Quick start 
curl -X POST https://api.jptranscribe.com/v1/transcribe \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -F "file=@recording.wav" \
  -F "language=ja" \
  -F "translate=true"
PoolCommunity worker CPU
ja → enTranscription + translation
RESTJSON responses, OpenAPI 1.3
$0Free forever

Features

Everything you need for Japanese audio

Accurate Japanese ASR with optional English output. No ML ops — send audio, get text.

🎙

Japanese transcription

Lightweight Vosk Japanese model on CPU — tuned for short clips, voice notes, and stream segments up to 90 seconds.

🌐

English translation

Append translate=true to get English captions alongside or instead of Japanese text.

⚡

Simple REST API

Upload a file or send a URL. JSON in, JSON out. Bearer token auth. Standard HTTP status codes.

📄

OpenAPI spec

Download our OpenAPI 3 schema and generate clients in Python, TypeScript, Go, or any language.

🔒

Secure by default

HTTPS only. API keys per project. Audio deleted after processing — not used for training.

📊

Interactive API docs

Try endpoints at Swagger UI or upload audio on Try it.

Use cases

Built for real workflows

Streaming

Live stream captions

Pipe audio chunks from OBS or your app into JP Transcribe. Display English overlays for international viewers.

Mobile

Voice notes & meetings

Transcribe Japanese voice memos in productivity apps. Export searchable text or translated summaries.

Automation

Podcast & video pipelines

Batch-process uploaded media via REST. Integrate with your CMS or automation pipeline.

Research

Interview transcription

Convert Japanese field recordings to text for analysis.

How it works

Three steps to captions

1. Get a free API key

One click on get-key. No email, no credit card. Run worker.py for full access.

2. Send audio

POST a WAV, MP3, M4A, or FLAC file up to 10 MB (90 sec with worker). Your worker transcribes locally; server demo for short clips.

3. Use the text

Response includes Japanese text and optional English translation.

Start transcribing Japanese audio

Generate a free key in seconds. Full docs and OpenAPI spec are live.

Get free API key View documentation

FAQ

Common questions

What audio formats are supported?

WAV, MP3, M4A, FLAC, and OGG. Files up to 10 MB, clips up to 90 seconds. Runs on lightweight CPU — no GPU required.

How accurate is Japanese transcription?

Vosk small Japanese model on CPU. Best for clear speech in short clips. Noisy or long audio may reduce accuracy.

Is there a streaming / WebSocket API?

REST batch API is available at launch. WebSocket streaming for sub-second live captions is on the roadmap — contact us if you need beta access.

Do you store my audio?

Audio is processed in memory and deleted after the request completes. We do not use customer audio to train models.

Do I need to run a worker?

For full transcription, yes — run worker.py on your PC (setup guide). Without a worker you get 3 free server demos/day (max 30s clips) or spend pool credits earned by helping others.

How is the API priced?

Free forever — no credit card. Translation is always server-side and free. Transcription uses your CPU via the community pool, with optional server demos. See pricing.