Skip to content
CallMissed
Book a demo

Explore CallMissed

Back to all models
Speech to Textvoice-agentsstreamingmultilingualindia

Microsoft Speech

by Microsoft · Released Undisclosed

Microsoft's real-time streaming speech-to-text model for voice agents, processed in India. 81 languages (148 locales) including Hindi, Bengali, Tamil, Telugu, Marathi, Gujarati, Kannada, Malayalam, Punjabi, Odia, Assamese, Urdu, Bhojpuri and Indian English. One language per call. Voice sessions only. $1.0417 per audio hour.

Speech to Text

Microsoft Speech

Powered by Microsoft · Streaming ASR (Microsoft)

Context Window

N/A

Parameters

Undisclosed

Max Output

N/A

Category

Speech to Text

Overview

Microsoft Speech is Microsoft's real-time streaming speech recognizer, and on CallMissed it runs inside voice agents and voice sessions only. Audio is processed in India.

It hears 81 languages across 148 locales, including Hindi, Bengali, Tamil, Telugu, Marathi, Gujarati, Kannada, Malayalam, Punjabi, Odia, Assamese, Urdu, Bhojpuri and Indian English. It recognises one language per call: it does not auto-detect or follow a caller who switches language, so set the agent's language to the one the caller speaks.

It has no turn detection of its own, so the agent's turn detection applies. Keyterms set on the agent (stt_keyterms) are applied as a phrase list. It is streaming-only: POST /v1/audio/transcriptions returns 400 for this model. For recorded audio use saaras:v4, ink-whisper or deepgram-nova-3.

Pricing

MetricPrice
Price /hour₹100.0000

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

  • 81 languages (148 locales), including 13 Indian languages and Indian English
  • Audio processed in India
  • One language per call — set the agent language to the caller's
  • Voice sessions only — not available for file transcription

Benchmarks

BenchmarkScore
Languages81
Price$1.0417
DeploymentStreaming only

Technical Details

  • One language per call: no auto-detect and no following a caller who switches language
  • No model-native turn detection — the agent's turn detection applies
  • Keyterms (stt_keyterms) are applied as a phrase list
  • Streaming only — the file-transcription endpoint returns 400 for this model
  • Selected per agent or voice session via stt_model

Strengths

  • Covers 13 Indian languages and Indian English in one model
  • Audio processed in India
  • Keyterms applied as a phrase list

Limitations

  • One language per call — it does not follow a caller who switches language
  • No built-in turn detection
  • Streaming only — not available for file transcription (use saaras:v4 or ink-whisper for recorded audio)
  • Higher per-hour price than the Indic-specialist and Whisper-family models

Use Cases

Indian-language voice agentsContact-centre automation in a single known languageLive call transcription

API Example

curl https://api.callmissed.com/v1/voice/sessions \
  -H "Authorization: Bearer cm_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "stt_model": "microsoft-speech-stt",
    "language": "hi-IN"
  }'

Endpoint: POST /v1/voice/sessions (streaming only) · Model ID: microsoft-speech-stt

Try Microsoft Speech now

Get 1000 free API credits on signup. No credit card required.

Start free Book a DemoRead docs