neha
0:000:06
नमस्ते, मैं नेहा बोल रही हूं Acme से। आपका order कल deliver होगा, बस address confirm कर लीजिए।
Rendered by Mira TTS · 2026-09-03
Models · Text to speech
Hindi text to speech for phone calls.
Four voices in eight Indian languages, with audio ready for the phone line.
- per 10 lakh characters over HTTP²
- ₹1,800
- a minute, streaming live
- ₹0.90
- Indian languages, and Hinglish
- 8
- for OpenAI’s speech API
- Drop-in
Hear it in each language.
Text120 characters
नमस्ते, मैं आशु बोल रहा हूँ। आपके कर्ज़ की किश्त अगले हफ़्ते जमा होनी है। आप चाहें तो मैं अभी जमा करने का तरीका बता दूँ।
48 kHz · MonoReady when you are
0:00 / 0:07
- voice
- Ashu
- language
- Hindi
- speech
- 0:07
- characters
- 120
Mira TTS, voice ashu, rendered 16 Oct 2026. One request, not edited.
Text149 characters
Hi, मैं नेहा बोल रही हूँ Acme Store से। आपका order dispatch हो गया है और Friday तक deliver हो जाएगा। Address में कोई change करना हो तो अभी बता दीजिए।
48 kHz · MonoReady when you are
0:00 / 0:09
- voice
- Neha
- language
- Hinglish
- speech
- 0:09
- characters
- 149
Mira TTS, voice neha, rendered 16 Oct 2026. One request, not edited.
Text152 characters
Hi, this is Sameer from Acme Logistics. You applied for the delivery partner role in Pune, and I have a few quick questions. Is now a good time to talk?
48 kHz · MonoReady when you are
0:00 / 0:08
- voice
- Sameer
- language
- English
- speech
- 0:08
- characters
- 152
Mira TTS, voice sameer, rendered 16 Oct 2026. One request, not edited.
Text117 characters
નમસ્તે, હું શ્રુતિ, દાંતના ક્લિનિકમાંથી બોલું છું. કાલે સવારે ડૉક્ટર સાથે તમારી મુલાકાત નક્કી છે. તમે સમયસર આવી શકશો?
48 kHz · MonoReady when you are
0:00 / 0:08
- voice
- Shruti
- language
- Gujarati
- speech
- 0:08
- characters
- 117
Mira TTS, voice shruti, rendered 17 Oct 2026. One request, not edited.
Text147 characters
வணக்கம், நான் ஆஷு பேசுகிறேன். உங்கள் கடனுக்கான தவணையை அடுத்த வாரம் கட்ட வேண்டும். நீங்கள் விரும்பினால், எப்படிக் கட்டுவது என்று இப்போதே சொல்கிறேன்.
48 kHz · MonoReady when you are
0:00 / 0:09
- voice
- Ashu
- language
- Tamil
- speech
- 0:09
- characters
- 147
Mira TTS, voice ashu, rendered 17 Oct 2026. One request, not edited.
Text133 characters
నమస్కారం, నేను శ్రుతి, పళ్ళ క్లినిక్ నుండి మాట్లాడుతున్నాను. రేపు ఉదయం డాక్టర్ గారితో మీకు అపాయింట్మెంట్ ఉంది. మీరు సమయానికి రాగలరా?
48 kHz · MonoReady when you are
0:00 / 0:08
- voice
- Shruti
- language
- Telugu
- speech
- 0:08
- characters
- 133
Mira TTS, voice shruti, rendered 17 Oct 2026. One request, not edited.
Listen
Hear two voices read the lines a business sends.
Original model output, not edited. The text under each player is what we sent.
ashu
0:000:06
नमस्ते, मैं आशु बोल रहा हूं। आपकी EMI पांच तारीख को due है — क्या online pay करेंगे?
Rendered by Mira TTS · 2026-09-03
Why Mira TTS
It is made for Indian phone calls.
Eight Indian languages, audio the phone line takes as it is, and an endpoint your code already knows.
-
01
Speaks the way India does
Hindi, Gujarati, Marathi, Bengali, Tamil, Telugu, Kannada and Malayalam. In Hinglish it switches from “नमस्ते” to “order confirm” in the same sentence.
-
02
Ready for the phone line
Ask for 8 kHz μ-law or A-law and send it straight down the line, or take 48 kHz PCM for everything else.
-
03
Drops into your stack
An OpenAI-compatible speech endpoint. Stream PCM as it renders, or ask for a WAV file.
Pricing
Text to speech is priced by the minute or by the character.
Start on pay as you go, and move to Business when you need more at once.
Text to speech
Mira TTS
₹0.90a minute connected, on the WebSocket²
₹1.80per 1,000 characters, over HTTP: ₹1,800 per 10 lakh (1 million)
- Monthly
- No commitment
- At once
- 5 concurrent requests
Pay as you go
- Monthly
- ₹50,000, credited to your wallet
- At once
- 25 concurrent requests
Business
- On the text-to-speech WebSocket you pay for time, not text: every second the socket is open is billed at the per-minute rate, and the speech it carries is not charged again by the character. HTTP requests are billed by the character.
- Shared endpoint · No contractual SLA. Top-ups use the same rates.
- Enterprise: Custom volume pricing, concurrency, dedicated endpoint and SLA.
- Audio
- 48 kHz mono PCM16 as a WAV file or raw PCM, or 8 kHz μ-law and A-law for phone lines.
- Voices
- ashu and sameer, neha and shruti: Hindi voices that code-switch into English, the same on every plan.
- Protocol
- OpenAI’s POST /v1/audio/speech, with model mira-tts. One sk_live_ key works for TTS and the v2 API.
- Languages
- Hindi, Gujarati, Marathi, Bengali, Tamil, Telugu, Kannada and Malayalam, and Hinglish, on the same endpoint.
Questions
Questions people ask about Mira TTS.
What does it cost?
₹1.80 per 1,000 characters over HTTP, which is ₹1,800 per 10 lakh characters, and only completed audio is charged. On the WebSocket you pay ₹0.90 for every minute the socket is open, metered by the second. All prices exclude GST.
Which languages and voices are available?
Hindi, Gujarati, Marathi, Bengali, Tamil, Telugu, Kannada and Malayalam are live, and so is Hinglish. The same four voices speak all of them: ashu, neha, shruti and sameer.
Can I send the audio straight to a phone call?
Yes. Ask for 8 kHz μ-law or A-law and it is ready for the line. For anything else, take 48 kHz mono PCM as a WAV file or raw PCM.
Which voice sounds better for my calls?
Listen on your own lines. Take twenty sentences from your calls, with names, rupee amounts and English words in them, and play them to people who speak the language every day. We don’t quote a leaderboard score for Mira TTS today.
How do I start?
Get an API key in the sandbox, then call the OpenAI-compatible speech endpoint with model mira-tts. The quickstart has working examples.
Still deciding? Ask us.
Fifteen minutes with the team that builds it. Bring your script, or your hardest call.
Book 15 minutes