About the Ursa voice
One of the newer built-in Grok voices usable in TTS and speech-to-speech sessions.
Ursa is one of the 23 voices xAI added when it expanded the Grok Voice API lineup, so you will not find it in the Grok app's voice picker; it is API-only. We suggest trying it first for nature documentaries, but every Grok voice handles every kind of text; the samples above let you judge in seconds.
Use Ursa in the xAI text to speech API
The voice id is ursa (case-insensitive). A complete request:
curl -X POST https://api.x.ai/v1/tts \
-H "Authorization: Bearer $XAI_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"text": "Hello, this is Ursa. [pause] Nice to meet you.",
"voice_id": "ursa",
"language": "en",
"speed": 1.0
}' \
--output ursa.mp3Add output_format for WAV, PCM or telephony codecs, with_timestamps for per-character timing, and the same voice_id as the voice of a Grok voice agent. Full parameter list in the Grok Voice API guide; official reference at docs.x.ai.
Languages Ursa can speak
English, Arabic (Egypt), Arabic (Saudi Arabia), Arabic (UAE), Bengali, Chinese (Simplified), French, German, Hindi, Indonesian, Italian, Japanese, Korean, Portuguese (Brazil), Portuguese (Portugal), Russian, Spanish (Mexico), Spanish (Spain), Turkish, Vietnamese. Pass the BCP-47 code (for example zh, ja, pt-BR) or auto.
Compare with the original Grok voices
- Grok Ara voice — narration, product demos
- Grok Eve voice — announcements, explainer videos
- Grok Leo voice — documentaries, audiobooks
- Grok Rex voice — ads, trailers, sports
- Grok Sal voice — podcasts, radio-style reads