Voices

Two production voices today, both trained on Hinglish: they read Devanagari Hindi, Latin-script English and code-mixed sentences in one breath, at 24 kHz.

Anjura anjuraWarm, clear, collections-ready. The default voice. Female.

Raj rajSteady, friendly, confident. Male.

Both samples read the same line, unedited engine output: नमस्ते, मैं सरल finserv से बोल रही हूँ. आपकी EMI ₹12,450 है, जो 5 नवंबर को due है. late fee से बचने के लिए आप UPI से pay कर सकते हैं.

Listing voices

curl -s https://api.vartalaap.si/v1/voices -H "x-api-key: $VARTALAAP_API_KEY"

{"voices":[{"name":"anjura","language":"hi-IN+en-IN","sample_rate":24000,"prompt_frames":102},
           {"name":"raj","language":"hi-IN+en-IN","sample_rate":24000,"prompt_frames":102}],
 "model":"pocket-tts-hinglish-v1"}

The list is filtered by the key's voice allow-list. Voice ids are lowercase and stable; new voices are announced in the changelog.

Choosing

	Anjura	Raj
Collections, reminders, KYC follow-ups	✓ default	✓
Order and delivery updates	✓	✓
Sales, onboarding	✓	✓ warmer close
Long-form reading (statements, policies)	✓	✓

Pace, pitch and loudness are not parameters of this model (speed other than 1.0 is rejected with 400); expressiveness comes from the text itself. See Writing text that sounds right. temperature (default 0.3) trades stability for variation; seed makes a request reproducible.

Your own voice

A bespoke voice on the same engine costs the same 20 paise a minute. Record about thirty minutes, we train and gate it, and it appears in your project under its own id. Request it under Train your voice in the portal; details in the guide.