Sarvam speechToText API
The speechToText API from Sarvam — 2 operation(s) for speechtotext.
OpenAPI Specification
openapi: 3.1.0
info:
title: Endpoints chat speechToText API
version: 1.0.0
servers:
- url: https://api.sarvam.ai
description: Production
tags:
- name: speechToText
paths:
/speech-to-text:
post:
operationId: transcribe
summary: Speech to Text
description: "## Speech to Text API\n\nThis API transcribes speech to text in multiple Indian languages and English. Supports transcription for interactive applications.\n\n### Available Options:\n- **REST API** (Current Endpoint): For quick responses under 30 seconds with immediate results\n- **Batch API**: For longer audio files, [Follow This Documentation](https://docs.sarvam.ai/api-reference-docs/api-guides-tutorials/speech-to-text/batch-api)\n - Supports diarization (speaker identification)\n\n### Note:\n- Pricing differs for REST and Batch APIs\n- Diarization is only available in Batch API with separate pricing\n- Please refer to [here](https://docs.sarvam.ai/api-reference-docs/pricing) for detailed pricing information"
tags:
- speechToText
parameters:
- name: api-subscription-key
in: header
required: true
schema:
type: string
responses:
'200':
description: Successful Response
content:
application/json:
schema:
$ref: '#/components/schemas/Sarvam_Model_API_SpeechToTextResponse'
'400':
description: Bad Request
content:
application/json:
schema:
$ref: '#/components/schemas/Sarvam_Model_API_ErrorMessage'
'403':
description: Forbidden
content:
application/json:
schema:
$ref: '#/components/schemas/Sarvam_Model_API_ErrorMessage'
'422':
description: Unprocessable Entity
content:
application/json:
schema:
$ref: '#/components/schemas/Sarvam_Model_API_ErrorMessage'
'429':
description: Quota Exceeded
content:
application/json:
schema:
$ref: '#/components/schemas/Sarvam_Model_API_ErrorMessage'
'500':
description: Internal Server Error
content:
application/json:
schema:
$ref: '#/components/schemas/Sarvam_Model_API_ErrorMessage'
'503':
description: Service Overloaded
content:
application/json:
schema:
$ref: '#/components/schemas/Sarvam_Model_API_ErrorMessage'
requestBody:
content:
multipart/form-data:
schema:
type: object
properties:
file:
type: string
format: binary
description: 'The audio file to transcribe. Supported formats include WAV, MP3, AAC, AIFF, OGG, OPUS, FLAC, MP4/M4A, AMR, WMA, WebM, and PCM formats. The API automatically detects most codec formats, but for PCM files (pcm_s16le, pcm_l16, pcm_raw), you must specify the input_audio_codec parameter. PCM files are supported only at 16kHz sample rate.
The API works best with audio files sampled at 16kHz. If the audio contains multiple channels, they will be merged into a single channel.'
model:
$ref: '#/components/schemas/Sarvam_Model_API_SpeechToTextModel'
description: 'Specifies the model to use for speech-to-text conversion.
- **saaras:v3** (default, recommended): State-of-the-art model with flexible output formats. Supports multiple modes via the `mode` parameter: transcribe, translate, verbatim, translit, codemix.
- **saarika:v2.5** (legacy): Transcribes audio in the spoken language. Kept for backward compatibility.'
mode:
oneOf:
- $ref: '#/components/schemas/Sarvam_Model_API_Mode'
- type: 'null'
default: transcribe
description: "Mode of operation. **Only applicable when using saaras:v3 model.**\n\nExample audio: 'मेरा फोन नंबर है 9840950950'\n\n- **transcribe** (default): Standard transcription in the original language with proper formatting and number normalization.\n - Output: `मेरा फोन नंबर है 9840950950`\n\n- **translate**: Translates speech from any supported Indic language to English.\n - Output: `My phone number is 9840950950`\n\n- **verbatim**: Exact word-for-word transcription without normalization, preserving filler words and spoken numbers as-is.\n - Output: `मेरा फोन नंबर है नौ आठ चार zero नौ पांच zero नौ पांच zero`\n\n- **translit**: Romanization - Transliterates speech to Latin/Roman script only.\n - Output: `mera phone number hai 9840950950`\n\n- **codemix**: Code-mixed text with English words in English and Indic words in native script.\n - Output: `मेरा phone number है 9840950950`"
language_code:
$ref: '#/components/schemas/Sarvam_Model_API_SpeechToTextLanguage'
description: 'Specifies the language of the input audio in BCP-47 format.
**Note:** This parameter is optional for `saarika:v2.5` model.
**Available Options:**
- `unknown`: Use when the language is not known; the API will auto-detect.
- `hi-IN`: Hindi
- `bn-IN`: Bengali
- `kn-IN`: Kannada
- `ml-IN`: Malayalam
- `mr-IN`: Marathi
- `od-IN`: Odia
- `pa-IN`: Punjabi
- `ta-IN`: Tamil
- `te-IN`: Telugu
- `en-IN`: English
- `gu-IN`: Gujarati
**Additional Options (saaras:v3 only):**
- `as-IN`: Assamese
- `ur-IN`: Urdu
- `ne-IN`: Nepali
- `kok-IN`: Konkani
- `ks-IN`: Kashmiri
- `sd-IN`: Sindhi
- `sa-IN`: Sanskrit
- `sat-IN`: Santali
- `mni-IN`: Manipuri
- `brx-IN`: Bodo
- `mai-IN`: Maithili
- `doi-IN`: Dogri'
input_audio_codec:
$ref: '#/components/schemas/Sarvam_Model_API_InputAudioCodec'
description: Input Audio codec/format of the input file. PCM files are supported only at 16kHz sample rate.
required:
- file
/speech-to-text-translate:
post:
operationId: translate
summary: Speech To Text Translate
description: "## Speech to Text Translation API\n\nThis API automatically detects the input language, transcribes the speech, and translates the text to English.\n\n### Available Options:\n- **REST API** (Current Endpoint): For quick responses under 30 seconds with immediate results\n- **Batch API**: For longer audio files [Follow this documentation](https://docs.sarvam.ai/api-reference-docs/api-guides-tutorials/speech-to-text/batch-api)\n - Supports diarization (speaker identification)\n\n### Note:\n- Pricing differs for REST and Batch APIs\n- Diarization is only available in Batch API with separate pricing\n- Please refer to [here](https://docs.sarvam.ai/api-reference-docs/pricing) for detailed pricing information"
tags:
- speechToText
parameters:
- name: api-subscription-key
in: header
required: true
schema:
type: string
responses:
'200':
description: Successful Response
content:
application/json:
schema:
$ref: '#/components/schemas/Sarvam_Model_API_SpeechToTextTranslateResponse'
'400':
description: Bad Request
content:
application/json:
schema:
$ref: '#/components/schemas/Sarvam_Model_API_ErrorMessage'
'403':
description: Forbidden
content:
application/json:
schema:
$ref: '#/components/schemas/Sarvam_Model_API_ErrorMessage'
'422':
description: Unprocessable Entity
content:
application/json:
schema:
$ref: '#/components/schemas/Sarvam_Model_API_ErrorMessage'
'429':
description: Quota Exceeded
content:
application/json:
schema:
$ref: '#/components/schemas/Sarvam_Model_API_ErrorMessage'
'500':
description: Internal Server Error
content:
application/json:
schema:
$ref: '#/components/schemas/Sarvam_Model_API_ErrorMessage'
'503':
description: Service Overloaded
content:
application/json:
schema:
$ref: '#/components/schemas/Sarvam_Model_API_ErrorMessage'
requestBody:
content:
multipart/form-data:
schema:
type: object
properties:
file:
type: string
format: binary
description: 'The audio file to transcribe. Supported formats include WAV, MP3, AAC, AIFF, OGG, OPUS, FLAC, MP4/M4A, AMR, WMA, WebM, and PCM formats. The API automatically detects most codec formats, but for PCM files (pcm_s16le, pcm_l16, pcm_raw), you must specify the input_audio_codec parameter. PCM files are supported only at 16kHz sample rate.
Works best at 16kHz. Multiple channels will be merged.'
prompt:
type:
- string
- 'null'
description: Conversation context can be passed as a prompt to boost model accuracy. However, the current system is at an experimentation stage and doesn't match the prompt performance of large language models.
model:
$ref: '#/components/schemas/Sarvam_Model_API_SpeechToTextTranslateModel'
description: "Model to be used for speech to text translation.\n\n- **saaras:v2.5** (default): Translation model that translates audio from any spoken Indic language to English.\n - Example: Hindi audio → English text output\n\nFor the latest model (saaras:v3), use the `/speech-to-text` endpoint with `mode=\"translate\"`."
input_audio_codec:
$ref: '#/components/schemas/Sarvam_Model_API_InputAudioCodec'
description: Audio codec/format of the input file. Our API automatically detects all codec formats, but for PCM files specifically (pcm_s16le, pcm_l16, pcm_raw), you must pass this parameter. PCM files are supported only at 16kHz sample rate.
required:
- file
components:
schemas:
Sarvam_Model_API_DiarizedEntry:
type: object
properties:
transcript:
type: string
description: transcript of the segment of that audio
start_time_seconds:
type: number
format: double
description: Start time of the word in seconds.
end_time_seconds:
type: number
format: double
description: End time of the word in seconds.
speaker_id:
type: string
description: Speaker ID for the word.
required:
- transcript
- start_time_seconds
- end_time_seconds
- speaker_id
title: Sarvam_Model_API_DiarizedEntry
Sarvam_Model_API_DiarizedTranscript:
type: object
properties:
entries:
type: array
items:
$ref: '#/components/schemas/Sarvam_Model_API_DiarizedEntry'
description: List of diarized transcript entries.
required:
- entries
title: Sarvam_Model_API_DiarizedTranscript
Sarvam_Model_API_SpeechToTextTranslateLanguage:
type: string
enum:
- hi-IN
- bn-IN
- kn-IN
- ml-IN
- mr-IN
- od-IN
- pa-IN
- ta-IN
- te-IN
- gu-IN
- en-IN
- as-IN
- ur-IN
- ne-IN
- kok-IN
- ks-IN
- sd-IN
- sa-IN
- sat-IN
- mni-IN
- brx-IN
- mai-IN
- doi-IN
description: 'Languages supported for Speech-to-Text-Translate (detected source language).
**saaras:v2.5 supports (11 languages):** hi-IN, bn-IN, kn-IN, ml-IN, mr-IN, od-IN, pa-IN, ta-IN, te-IN, gu-IN, en-IN
For additional languages (as-IN, ur-IN, ne-IN, kok-IN, ks-IN, sd-IN, sa-IN, sat-IN, mni-IN, brx-IN, mai-IN, doi-IN), use saaras:v3 via the `/speech-to-text` endpoint with `mode="translate"`.'
title: Sarvam_Model_API_SpeechToTextTranslateLanguage
Sarvam_Model_API_TimestampsModel:
type: object
properties:
words:
type: array
items:
type: string
description: List of words in the transcript.
start_time_seconds:
type: array
items:
type: number
format: double
description: List of start times of words in seconds.
end_time_seconds:
type: array
items:
type: number
format: double
description: List of end times of words in seconds.
required:
- words
- start_time_seconds
- end_time_seconds
title: Sarvam_Model_API_TimestampsModel
Sarvam_Model_API_SpeechToTextLanguage:
type: string
enum:
- unknown
- hi-IN
- bn-IN
- kn-IN
- ml-IN
- mr-IN
- od-IN
- pa-IN
- ta-IN
- te-IN
- en-IN
- gu-IN
- as-IN
- ur-IN
- ne-IN
- kok-IN
- ks-IN
- sd-IN
- sa-IN
- sat-IN
- mni-IN
- brx-IN
- mai-IN
- doi-IN
description: 'Languages supported for Speech-to-Text.
**saarika:v2.5 supports (12 languages):** unknown, hi-IN, bn-IN, kn-IN, ml-IN, mr-IN, od-IN, pa-IN, ta-IN, te-IN, en-IN, gu-IN
**saaras:v3 supports all 23 languages** including: as-IN, ur-IN, ne-IN, kok-IN, ks-IN, sd-IN, sa-IN, sat-IN, mni-IN, brx-IN, mai-IN, doi-IN'
title: Sarvam_Model_API_SpeechToTextLanguage
Sarvam_Model_API_SpeechToTextTranslateResponse:
type: object
properties:
request_id:
type:
- string
- 'null'
transcript:
type: string
description: Transcript of the provided speech
language_code:
oneOf:
- $ref: '#/components/schemas/Sarvam_Model_API_SpeechToTextTranslateLanguage'
- type: 'null'
description: This will return the BCP-47 code of language spoken in the input. If multiple languages are detected, this will return language code of most predominant spoken language. If no language is detected, this will be null
diarized_transcript:
oneOf:
- $ref: '#/components/schemas/Sarvam_Model_API_DiarizedTranscript'
- type: 'null'
description: Diarized transcript of the provided speech
language_probability:
type:
- number
- 'null'
format: double
description: 'Float value (0.0 to 1.0) indicating the probability of the detected language being correct. Higher values indicate higher confidence.
**When it returns a value:**
- When `language_code` is not provided in the request
- When `language_code` is set to `unknown`
**When it returns null:**
- When a specific `language_code` is provided (language detection is skipped)
The parameter is always present in the response.'
required:
- request_id
- transcript
- language_code
title: Sarvam_Model_API_SpeechToTextTranslateResponse
Sarvam_Model_API_SpeechToTextResponse:
type: object
properties:
request_id:
type:
- string
- 'null'
transcript:
type: string
description: The transcribed text from the provided audio file.
timestamps:
oneOf:
- $ref: '#/components/schemas/Sarvam_Model_API_TimestampsModel'
- type: 'null'
description: Contains timestamps for the transcribed text. This field is included only if with_timestamps is set to true
diarized_transcript:
oneOf:
- $ref: '#/components/schemas/Sarvam_Model_API_DiarizedTranscript'
- type: 'null'
description: Diarized transcript of the provided speech
language_code:
type:
- string
- 'null'
description: This will return the BCP-47 code of language spoken in the input. If multiple languages are detected, this will return language code of most predominant spoken language. If no language is detected, this will be null
language_probability:
type:
- number
- 'null'
format: double
description: 'Float value (0.0 to 1.0) indicating the probability of the detected language being correct. Higher values indicate higher confidence.
**When it returns a value:**
- When `language_code` is not provided in the request
- When `language_code` is set to `unknown`
**When it returns null:**
- When a specific `language_code` is provided (language detection is skipped)
The parameter is always present in the response.'
required:
- request_id
- transcript
- language_code
title: Sarvam_Model_API_SpeechToTextResponse
Sarvam_Model_API_Mode:
type: string
enum:
- transcribe
- translate
- verbatim
- translit
- codemix
description: "Mode of operation for saaras:v3 model.\n\nExample audio: 'मेरा फोन नंबर है 9840950950'\n\n- **transcribe** (default): Standard transcription in the original language with proper formatting and number normalization.\n - Output: `मेरा फोन नंबर है 9840950950`\n\n- **translate**: Translates speech from any supported Indic language to English.\n - Output: `My phone number is 9840950950`\n\n- **verbatim**: Exact word-for-word transcription without normalization, preserving filler words and spoken numbers as-is.\n - Output: `मेरा फोन नंबर है नौ आठ चार zero नौ पांच zero नौ पांच zero`\n\n- **translit**: Romanization - Transliterates speech to Latin/Roman script only.\n - Output: `mera phone number hai 9840950950`\n\n- **codemix**: Code-mixed text with English words in English and Indic words in native script.\n - Output: `मेरा phone number है 9840950950`"
title: Sarvam_Model_API_Mode
Sarvam_Model_API_SpeechToTextTranslateModel:
type: string
enum:
- saaras:v2.5
description: "Model to be used for speech to text translation.\n\n- **saaras:v2.5** (default): Translation model that translates audio from any spoken Indic language to English.\n - Example: Hindi audio → English text output\n\nFor the latest model (saaras:v3), use the `/speech-to-text` endpoint with `mode=\"translate\"`."
title: Sarvam_Model_API_SpeechToTextTranslateModel
Sarvam_Model_API_ErrorCode:
type: string
enum:
- invalid_request_error
- internal_server_error
- unprocessable_entity_error
- insufficient_quota_error
- invalid_api_key_error
- authentication_error
- not_found_error
- rate_limit_exceeded_error
title: Sarvam_Model_API_ErrorCode
Sarvam_Model_API_ErrorMessage:
type: object
properties:
error:
$ref: '#/components/schemas/Sarvam_Model_API_ErrorDetails'
description: Error details
required:
- error
title: Sarvam_Model_API_ErrorMessage
Sarvam_Model_API_InputAudioCodec:
type: string
enum:
- wav
- x-wav
- wave
- mp3
- mpeg
- mpeg3
- x-mp3
- x-mpeg-3
- aac
- x-aac
- aiff
- x-aiff
- ogg
- opus
- flac
- x-flac
- mp4
- x-m4a
- amr
- x-ms-wma
- webm
- pcm_s16le
- pcm_l16
- pcm_raw
description: Audio codec/format of the input file. Our API automatically detects all codec formats, but for PCM files specifically (pcm_s16le, pcm_l16, pcm_raw), you must pass this parameter. PCM files are supported only at 16kHz sample rate.
title: Sarvam_Model_API_InputAudioCodec
Sarvam_Model_API_SpeechToTextModel:
type: string
enum:
- saaras:v3
- saarika:v2.5
description: 'Model to be used for speech to text.
- **saaras:v3** (default, recommended): State-of-the-art model with flexible output formats. Supports multiple modes via the `mode` parameter: transcribe, translate, verbatim, translit, codemix.
- **saarika:v2.5** (legacy): Transcribes audio in the spoken language. Kept for backward compatibility.'
title: Sarvam_Model_API_SpeechToTextModel
Sarvam_Model_API_ErrorDetails:
type: object
properties:
request_id:
type:
- string
- 'null'
message:
type: string
description: Message describing the error
code:
$ref: '#/components/schemas/Sarvam_Model_API_ErrorCode'
description: Error code for the specific error that has occurred. Refer to the error code documentation for more details.
required:
- request_id
- message
- code
title: Sarvam_Model_API_ErrorDetails
securitySchemes:
ApiKeyAuth:
type: apiKey
in: header
name: api-subscription-key
HTTPBearer:
type: http
scheme: bearer