Generate audio from text
Kinyarwanda is unlimited and free in the studio. English and French use tokens by length.
Text to Speech
Pick a voice, paste your script, generate.
Record 6–12 seconds of clear speech in English or French.
Max 20MB. Your file is deleted immediately after processing.
Your text is not stored. Audio is generated on demand and not retained.
Kinyarwanda uses IjwiLab FlexKIN. English and French use Piper voices.
Voice cloning retired.
/admin/ — Edit pronunciation rules
Recent generations
Why IjwiLab
Built for Rwanda first — ready for production voice in multilingual AI systems.
Native Kinyarwanda
Pronunciation and cadence built for the language — not bolted on.
Real voices
Named speakers with distinct tone — pick the right presence for your script.
Voice API
One endpoint for rw, en, and fr — keys, tokens, and docs for AI receptionists.
Ship voice into your product
Create an API key, top up tokens, and call POST /api/v1/tts.php.
IjwiLab