Microsoft Neural voices
by Microsoft · Released Undisclosed
Microsoft's neural text-to-speech, processed in India. 57 voices across 13 languages including Hindi, Bengali, Tamil, Telugu, Marathi, Gujarati, Kannada, Malayalam, Punjabi, Odia, Assamese and Urdu. Plain text, adjustable speaking speed. Voice sessions only. $0.1563 per 10K characters.
Microsoft Neural voices
Powered by Microsoft · Neural TTS (Microsoft)
Context Window
N/A
Parameters
Undisclosed
Max Output
N/A
Category
Text to Speech
Overview
Microsoft Neural voices is Microsoft's neural text-to-speech service, and on CallMissed it runs inside voice agents and voice sessions only. Audio is processed in India.
There are 57 voices across 13 languages: Assamese, Bengali, English, Gujarati, Hindi, Kannada, Malayalam, Marathi, Odia, Punjabi, Tamil, Telugu and Urdu. A voice speaks only its own locale, except the four Indic voices, which speak English (India) plus 11 Indian languages. The default voice is en-in-neerja. Gender is the value Microsoft publishes for each voice.
It takes plain text, with no SSML, audio tags or delivery tags, and it normalises rupee amounts and dates itself. The one control is speaking speed (tts_speed, 0.5 to 2.0, default 1.0). POST /v1/audio/speech returns an error for it.
Pricing
| Metric | Price |
|---|---|
| Price /10K chars | ₹15.0000 |
1 credit = ₹1 ≈ US$0.0104 (US$1 = ₹96). Transparent per-call pricing — you pay only for what you call, with no seat fees or minimums.
Key Highlights
- 57 voices across 13 languages, including 12 Indian languages
- Audio processed in India
- Four Indic voices speak English plus 11 Indian languages
- Voice sessions only — not available on the speech endpoint
Benchmarks
| Benchmark | Score |
|---|---|
| Languages | 13 |
| Voices | 57 |
| Price | $0.1563 |
| Deployment | Voice sessions only |
Technical Details
- Languages: as, bn, en, gu, hi, kn, ml, mr, or, pa, ta, te, ur
- A voice speaks only its own locale, except the four Indic voices (English plus 11 Indian languages)
- Plain text only — no SSML, audio tags or delivery tags
- Speaking speed (tts_speed) from 0.5 to 2.0, default 1.0
- Normalises rupee amounts and dates itself
- Voice sessions only — POST /v1/audio/speech returns 400 for this model
Strengths
- Native voices for 12 Indian languages plus English
- Audio processed in India
- Handles rupee amounts and dates without preprocessing
Limitations
- Plain text only — no SSML, audio tags or delivery tags
- Speaking speed is the only control
- Voice sessions only — not available on the speech endpoint
Use Cases
API Example
curl https://api.callmissed.com/v1/voice/sessions \
-H "Authorization: Bearer cm_YOUR_KEY" \
-H "Content-Type: application/json" \
-d '{
"tts_model": "microsoft-neural-tts",
"voice": "hi-in-swara",
"language": "hi"
}'Endpoint: POST /v1/voice/sessions (voice sessions only) · Model ID: microsoft-neural-tts
Try Microsoft Neural voices now
Get 1000 free API credits on signup. No credit card required.