Mira TTS
₹1.80 per 1,000 characters.⁴
And ₹15 an hour to transcribe a recorded call. Both from one rupee wallet.
Azure Speech alternative
Voices that switch into English mid-sentence and transcripts that keep each word in its own script, on one rupee wallet.
| Azure Speech | Voice Infra |
|---|---|
| Text to speech price | |
| ₹1,424per 10 lakh characters, neural voices ($15) | ₹1,800per 10 lakh characters |
| Speech to text price | |
| ₹17.09an hour, batch ($0.18) | ₹15an hour, recorded calls |
| Indian languages | |
| Hindi and eight more | Hindi and seven more, plus Hinglish |
| Phone-line audio | |
| 8 kHz μ-law and A-law | 8 kHz μ-law and A-law |
| Speaker labels | |
| Extra on real-time | Included on recorded calls |
Side by side
| Option | Bar | ₹ per 10 lakh characters |
|---|---|---|
| ₹1,424$15 | ||
| ₹2,089$22 | ||
| ₹1,800 |
Dollar prices converted at ₹94.96, the ECB reference rate on 16 October 2026. The rate.
10 lakh is 1 million. The first half-million neural characters each month are free.
From azure.microsoft.com/en-us/pricing/details/speech, checked . Ours from our price list.
| Option | Bar | ₹ per hour of audio |
|---|---|---|
| ₹94.96$1 | ||
| ₹34.19$0.36 | ||
| ₹17.09$0.18 | ||
| ₹15 |
Dollar prices converted at ₹94.96, the ECB reference rate on 16 October 2026. The rate.
From azure.microsoft.com/en-us/pricing/details/speech, checked . Ours from our price list.
| Detail | Azure Speech | Voice Infra |
|---|---|---|
| Text to speech price | Neural and Neural HD Flash: $15 per 1M characters Neural HD and MAI-Voice: $22 per 1M characters azure.microsoft.com | ₹1.80 per 1,000 characters, on pay as you go and on Business (₹50,000/month commitment), before GST |
| Speech to text price | Real-time $1 per hour fast transcription $0.36; batch $0.18; speaker diarization on real-time $0.3 per hour per feature azure.microsoft.com | Recorded calls: ₹15 per hour of audio (₹0.25 a minute), speaker labels included, charged once when the job completes. Live recognition is ₹0.30 a minute of audio over HTTP or a WebSocket |
To speak, Azure lists ₹1,424 per million characters on neural voices and ₹2,089 on Neural HD. Ours is ₹1,800 per million, and a request that fails costs nothing.
To transcribe, Azure’s batch rate is ₹17.09 an hour, fast transcription ₹34.19 and live ₹94.96. Ours is ₹15 an hour of recorded audio, which is 25 paise a minute, with speaker labels included. We charge once, when the job is done. A job that fails or comes back degraded is free. Azure’s free tier, 5 hours and half a million characters a month, is a nice way to prototype.
On 499 public Hindi phone recordings at 8 kHz, Mira STT made 18.3% word errors, against 21.6% for Sarvam’s Saaras v4 on the same clips. On clean read speech Saaras does better, 5.8% against our 10.1%. Azure wasn’t in that test, so it tells you nothing about Azure. The fair test is twenty of your own calls, sent to both.
| Detail | Azure Speech | Voice Infra |
|---|---|---|
| Indian languages | hi-IN neural voices (Swara, Madhur, Aarav, Ananya, Kavya and others) and MAI-Voice voices in preview two neural voices each for Gujarati, Tamil, Telugu, Bengali, Marathi, Kannada, Malayalam and Punjabi learn.microsoft.com | Hindi, Gujarati, Marathi, Bengali, Tamil, Telugu, Kannada and Malayalam, plus Hinglish, in speech recognition and voice |
| Detail | Azure Speech | Voice Infra |
|---|---|---|
| Where it runs | Central India supports neural TTS and real-time, fast and batch transcription South India is not supported for standalone speech learn.microsoft.com | Speech recognition and the model run on our own GPUs in India |
| Detail | Azure Speech | Voice Infra |
|---|---|---|
| How to switch | Azure's Speech SDK and REST APIs, with an endpoint per region learn.microsoft.com | TTS on OpenAI's POST /v1/audio/speech transcription on POST /v2/transcriptions; one sk_live_ key for both |
| Try it free | Free F0 tier: 0.5 million neural characters and 5 audio hours of real-time transcription a month azure.microsoft.com | No sign-up to look around the sandbox ₹200 of credit once you confirm a work email; clips you play in the console are not charged |
Spot something out of date? Tell us. Every platform in one table: voice AI platforms compared.
Why switch
Everything a phone call in India needs, from one team, priced in rupees.
₹1.80 per 1,000 characters.⁴
And ₹15 an hour to transcribe a recorded call. Both from one rupee wallet.
Hindi, Gujarati, Marathi, Bengali, Tamil, Telugu, Kannada and Malayalam. And the Hinglish in between, mid-sentence.
The listening and the thinking: Mira STT and Gemma 4 26B. Mira TTS speaks.
Ask for 8 kHz μ-law or A-law and send it straight down the line. Or take 48 kHz PCM for everything else.
When you want the whole call, the same key and wallet run our voice agents.
AI agents: ₹5/min to start · ₹1.50/min all-in at volume.¹
Real lines from Mira TTS, one request each, unedited. Pick a language and press play.
Hear itBroad, or close to the call
Azure Speech is a broad, safe choice, with many Indian languages and a region in Central India. We're narrower and closer to the phone call. Our voices switch into English inside a sentence, our transcripts keep each word in its own script, and both bill one rupee wallet: ₹1.80 per 1,000 characters to speak, ₹15 an hour to transcribe.
Azure Speech lives inside Microsoft’s cloud and does a great deal. Its Hindi voices include Swara and Madhur, plus the newer MAI-Voice voices in preview. It transcribes Hindi live, in batch and in a fast mode, and it does all of that from Central India. It also speaks Bengali, Marathi, Kannada and more, which we don’t yet.
We do two things, both for calls in India. Mira TTS speaks Hindi, Gujarati, Marathi, Bengali, Tamil, Telugu, Kannada and Malayalam, and the mix in between. Mira Transcribe turns a recorded call into a transcript that says who spoke, with each word in its own script, just as people switched. One key and one rupee wallet cover both, and transcription runs on our GPUs in India.
Real lines from Mira TTS. Pick one, press play.
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
Mira TTS, voice neha, rendered 16 Oct 2026. One request, not edited.
Text119 characters
नमस्ते, मैं श्रुति बोल रही हूँ, दाँतों के क्लिनिक से। कल सुबह डॉक्टर साहब से आपका मिलना तय है। क्या आप समय पर आ पाएँगे?
48 kHz · MonoReady when you are
0:00 / 0:08
Mira TTS, voice shruti, rendered 16 Oct 2026. One request, not edited.
Text117 characters
નમસ્તે, હું આશુ બોલું છું. તમારી લોનનો હપ્તો આવતા અઠવાડિયે ભરવાનો છે. તમે કહો તો હું અત્યારે જ ભરવાની રીત સમજાવી દઉં.
48 kHz · MonoReady when you are
0:00 / 0:08
Mira TTS, voice ashu, rendered 17 Oct 2026. One request, not edited.
Text167 characters
வணக்கம், நான் நேஹா பேசுகிறேன். உங்கள் பார்சல் நாளை மாலைக்குள் வந்து சேரும். அந்த நேரத்தில் நீங்கள் வீட்டில் இல்லையென்றால் சொல்லுங்கள், வேறு நாளுக்கு மாற்றி விடுகிறோம்.
48 kHz · MonoReady when you are
0:00 / 0:10
Mira TTS, voice neha, rendered 17 Oct 2026. One request, not edited.
Text114 characters
నమస్కారం, నేను సమీర్ మాట్లాడుతున్నాను. నిన్న మీరు మా కొత్త ఇళ్ళ గురించి అడిగారు. ఇప్పుడు కొంచెం సేపు మాట్లాడవచ్చా?
48 kHz · MonoReady when you are
0:00 / 0:08
Mira TTS, voice sameer, rendered 17 Oct 2026. One request, not edited.
Which to pick
Sometimes it's them. We'll say so.
Moving
Your code stays. Three values change.
Step 1
Look around with no sign-up. When you are ready, create a key in Developers.
Step 2
The base URL, the model and the voice. The rest of your OpenAI code stays as it is.
Step 3
Four voices, each speaking Hindi, Gujarati, Marathi, Bengali, Tamil, Telugu, Kannada and Malayalam.
For speech, swap the Speech SDK call for OpenAI’s speech API: POST /v1/audio/speech, with model mira-tts, a voice and a format. Ask for ulaw_8000 or alaw_8000 when the audio goes to a phone line.
For recordings, upload the file to POST /v2/transcriptions with a line of context and any names you want spelled right. Then check the job until it’s done. A recording can run to two hours. The transcription docs show both steps.
Building a calling agent on Azure? Ours already listen, think and speak in one place, for ₹1.50 a minute all-in on the volume plan.
Python
import os
from openai import OpenAI
client = OpenAI(
api_key=os.environ["MIRA_API_KEY"],
base_url="https://sandbox.voice.miraiminds.co/v1", # 1. our URL
)
with client.audio.speech.with_streaming_response.create(
model="mira-tts", # 2. our model
voice="ashu", # 3. one of our voices
response_format="pcm", # always send it
input="चलिए, आपका order कल deliver होगा।",
) as response:
for chunk in response.iter_bytes():
... # raw 16-bit mono PCM at 48 kHz, playable as it arrives cURL
curl --fail-with-body \
https://sandbox.voice.miraiminds.co/v1/audio/speech \
-H "Authorization: Bearer $MIRA_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "mira-tts",
"voice": "neha",
"response_format": "wav",
"input": "नमस्ते, आपका ऑर्डर तैयार है।"
}' \
--output speech.wav Questions
At the ECB rate, Azure's neural voices come to ₹1,424 ($15) per million characters, and Neural HD to ₹2,089 ($22). We charge ₹1,800. Azure's first half-million characters each month are free.
Azure's batch rate is ₹17.09 ($0.18) an hour, a little above our ₹15. Fast transcription is ₹34.19 and real-time ₹94.96. Our ₹15 includes speaker labels; on Azure's real-time service they cost extra.
Inside our voice agents, yes, on every call. As a standalone API, live transcription is in private preview, so ask us for access. Recorded calls work through the public API today.
Our docs cover plain text, up to 2,000 characters per request. If you rely on SSML tags or speaking styles, Azure has them and we don't document them.
Fifteen minutes with the team that builds it. Bring your script, or your hardest call.
Book 15 minutes