Kurdish voice AI, for developers
A fast, simple REST API for Kurdish text-to-speech and speech-to-text. 664 voices across Sorani and Kurmanji, word-level timing, and streaming. Free to start, no credit card.
# Text to speech → saves hello.wav
curl -X POST https://www.kurdishtts.com/api/tts-proxy \
-H "x-api-key: $KURDISH_TTS_KEY" \
-H "Content-Type: application/json" \
-d '{"speaker_id": "sorani_1", "model_version": "v4", "text": "سڵاو، ئەمە دەنگی کوردییە."}' \
--output hello.wav
POST /api/tts-proxy → returns audio/wav
Start in three steps
From zero to your first Kurdish audio in a couple of minutes.
Create an API key
Generate a key under Settings → API. It's shown once, so store it as an environment variable.
Call the endpoint
POST to /api/tts-proxy with your x-api-key header and a JSON body. Omit model_version for v3, or pass "v4" for the latest model.
Receive audio
The response is audio/wav. Add "include_timestamps": true for word-level timings and a JSON payload.
One API, three endpoints
Authenticate every request with your x-api-key header.
/api/tts-proxyText to speech
Send text, get natural Kurdish audio (WAV). 664 voices, v3 + v4 models, optional word-level timestamps.
/api/stt-proxySpeech to text
Upload an audio file, get a transcript back. Sorani or Kurmanji, up to 100MB per file.
/api/stt-stream-connectStreaming STT
Open a short-lived WebSocket and transcribe audio in real time as it arrives.
# Speech to text → JSON transcript
curl -X POST https://www.kurdishtts.com/api/stt-proxy \
-H "x-api-key: $KURDISH_STT_KEY" \
-F "file=@audio.wav" \
-F "dialect=sorani"
Pricing
Flat monthly plans, with one-time character packs for TTS. Billed per character for TTS, per minute of audio for STT.
Free
- 20,000 Characters per month
- Sorani and Kurmanji dialects
- 4 voices (1 male + 1 female per dialect)
Just Starting Out (v3)
- 1,000,000 Characters per month
- Sorani and Kurmanji dialects
- All 198 v3 voices (Full voice library)
TTS v4 Starter
- TTS v4 model, superior quality
- 1,000,000 characters per month
- All 664 v4 voices
- Both Sorani & Kurmanji dialects
- Download audio files
Kurdish API Pro
TTS + STT BundleIncludes everything in both TTS and STT plans
TTS Plan Comparison
| Feature | Free | Just Starting Out (v3) | TTS v4 Starter | Kurdish API Pro |
|---|---|---|---|---|
| TTS Model | v3 | v3 | v4 | v3 + v4 |
| Characters (per month) | 20,000 | 1,000,000 | 1,000,000 | 1,000,000 |
| Dialects | Sorani & Kurmanji | Sorani & Kurmanji | Sorani & Kurmanji | Sorani & Kurmanji |
| Available Voices | 4 voices | 198 voices | 664 voices | All voices (v3 + v4) |
| STT Audio | 100 hrs / month |
Need more credits?
Top up your account instantly. Credits never expire and stack on top of your subscription.
500K Characters
$0.006 / 1K chars
1M Characters
$0.005 / 1K chars
5M Characters
$0.004 / 1K chars
Limits & quotas
Honest, published limits. Free-tier and TTS requests past quota return a 403 with an upgrade hint, never a silent failure. Paid STT plans keep transcribing past their included minutes — we email you to upgrade.
- Per-request text
- 500 chars (free) · 5,000 (paid)
- STT upload size
- 10 · 50 · 100 MB by plan
- TTS quota
- 20k (free) · 1M chars / mo (paid)
- STT audio
- 2 · 20 · 100 hrs / mo (120 · 1,200 · 6,000 min)
- Streaming sessions
- 20 · 100 · 500 / mo
- Max stream length
- 2 · 10 · 30 min
- Rate limiting
- Quota-based (403 past quota)
- Dialect
- Inferred from speaker_id prefix
What it costs
Simple, transparent unit prices:
A 500-word article ≈ 3,000 characters ≈ $0.015
≈ $0.0042 per minute of audio — a 5-min clip ≈ $0.02
Build with Kurdish voice today
Generate a key and make your first call in minutes. Free tier, no credit card.