Developer platform
Playground
Talk to it in Derja. Or French. Or both.
curl https://baarcha.tn/api/v1/chat/completions \
-N \
-H "Authorization: Bearer $BAARCHA_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "hannibal",
"messages": [
{
"role": "user",
"content": "شكونك إنتي بالضبط؟"
}
],
"temperature": 0.7,
"max_tokens": 512,
"stream": true
}'Runs against the public v1 API with your key. This page executes the same call keylessly as a demo.
Say something in Derja and hear it come back as text. No microphone handy? The twelve sample clips below are the fallback.
Or fall back to a sample clip
All twelve are Dido's own renders: this demo feeds our text-to-speech straight back into our ears. That makes them clean, labelled audio for wiring up a request, and weak evidence about the model. The microphone above is the test that counts, because it is audio nothing here produced.
Each clip's known transcript stays hidden until you transcribe it, then appears next to what came back. An answer printed above the question is not a demo.
Transcript hidden until you transcribe it
Transcript hidden until you transcribe it
Transcript hidden until you transcribe it
Transcript hidden until you transcribe it
Transcript hidden until you transcribe it
Transcript hidden until you transcribe it
Transcript hidden until you transcribe it
Transcript hidden until you transcribe it
Transcript hidden until you transcribe it
Transcript hidden until you transcribe it
Transcript hidden until you transcribe it
Transcript hidden until you transcribe it
curl https://baarcha.tn/api/v1/audio/transcriptions \
-H "Authorization: Bearer $BAARCHA_API_KEY" \
-F file=@'female-street.mp3' \
-F model=fennecRuns against the public v1 API with your key. This page executes the same call keylessly as a demo.
curl https://baarcha.tn/api/v1/audio/speech \
-H "Authorization: Bearer $BAARCHA_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "dido",
"input": "الشارع مسكر اليوم على خاطر الأشغال، لازم نعملو دورة طويلة باش نوصلو للخدمة في الوقت.",
"voice": "female"
}' \
--output speech.mp3Runs against the public v1 API with your key. This page executes the same call keylessly as a demo.
curl https://baarcha.tn/api/v1/voices \
-H "Authorization: Bearer $BAARCHA_API_KEY" \
-F name='صوتي' \
-F consent=true \
-F sample=@'take-one.wav'Answers with a vc_… id already ready: one clip in, a usable voice out, in a second or two. Pass the id as voice above; there is nothing to poll. GET /api/v1/voices/{id} and the status field are still part of the contract and still worth branching on, because they are what lets a slower build appear later without breaking a client that already reads them. consent is required by the server on both dialects, and a key does not skip what the dashboard asks for. Full reference in the voices docs.