Speech to text transcription

Convert audio to text using advanced speech recognition.

Complete File Upload (Standard) Use Content-Type: multipart/form-data to upload the complete audio file in one request. Maximum file size: 25MB.

Example:

curl -X POST "http://localhost:3000/api/v1/llm/speech/transcriptions?language=en" \
  -F "file=@audio.flac"

Chunked Upload (Streaming) Use Transfer-Encoding: chunked header to stream audio data in chunks as it’s being recorded. No need to know total file size upfront. Server buffers chunks until complete before processing. Maximum total size: 25MB.

Example:

curl -X POST "http://localhost:3000/api/v1/llm/speech/transcriptions?language=en" \
  -H "Transfer-Encoding: chunked" \
  -H "Content-Type: multipart/form-data" \
  --data-binary @audio.flac

Supported Formats: FLAC, MP3, MP4, MPEG, MPGA, M4A, OGG, WAV, WebM

Query Parameters:

• model (optional): Transcription model identifier. Defaults to ‘auto’.
• language (optional): ISO-639-1 or BCP-47 language code (e.g., “en”, “en-US”). Auto-detects if not specified.
• prompt (optional): Legacy prompt parameter retained for backward compatibility.
• temperature (optional): Legacy temperature parameter retained for backward compatibility.
• include_speaker_data (optional): When true, include speaker diarization data and require WAV/PCM input. Otherwise transcription uses the standard compatibility path.

Response: Returns transcribed text in JSON format.

POST /api/v1/llm/speech/transcriptions

Python (SDK)

Python

from mka1 import SDK


with SDK(
    bearer_auth="<YOUR_BEARER_TOKEN_HERE>",
) as sdk:

    res = sdk.llm.speech.transcribe(file={
        "file_name": "example.file",
        "content": open("example.file", "rb"),
    }, model="auto", language="en", include_speaker_data=True, prompt="This is a technical podcast about machine learning.", temperature=0.2)

    # Handle response
    print(res)

200

{
  "text": "Hello, this is a sample transcription of the audio file.",
  "language": "en",
  "confidence": 0.95,
  "speakers": [
    {
      "speaker": "Speaker-1",
      "text": "Hello, this is a sample transcription of the audio file.",
      "confidence": 0.95,
      "offset_ms": 0,
      "duration_ms": 2100
    }
  ]
}

Authorizations

Authorization · string · header · required

Gateway auth: send Authorization: Bearer <mka1-api-key>. For multi-user server-side integrations, you can also send X-On-Behalf-Of: <external-user-id>.

Headers

X-On-Behalf-Of · string

Optional external end-user identifier forwarded by the API gateway.

Query Parameters

model · string · default: auto

Transcription model identifier. Defaults to 'auto' which selects the best available model. Ignored when speaker diarization is requested.

Example:

"auto"

language · string

The language of the input audio in ISO-639-1 or BCP-47 format (for example 'en' or 'en-US'). If not specified, the transcription service auto-detects the language.

Example:

"en"

include_speaker_data · boolean · default: false

Whether to include speaker-segment data. Defaults to false. When true, the response includes a speakers array split by detected speaker.

Example:

true

prompt · string

Legacy prompt parameter retained for backward compatibility. Used only by the fallback path for non-WAV uploads.

Example:

"This is a technical podcast about machine learning."

temperature · number

Legacy temperature parameter retained for backward compatibility. Used only by the fallback path for non-WAV uploads.

Required range: 0 <= x <= 1

Example:

0.2

Body

multipart/form-data

file · file · required

Audio file to transcribe.

File Requirements:

• Maximum size: 25MB
• Supported formats: FLAC, MP3, MP4, MPEG, MPGA, M4A, OGG, WAV, WebM
• Speaker diarization currently requires WAV / PCM input

Upload Options:

1. Standard: Upload complete file using multipart/form-data
2. Chunked: Stream file chunks using Transfer-Encoding: chunked header (useful for real-time recording)

Note: For chunked uploads, server buffers all chunks before processing. Transcription begins only after the complete file is received.

Response

200 - application/json

OK

Response from the transcription endpoint containing the transcribed text, detected language, confidence score, and optional speaker segments.

text · string · required

The transcribed text from the audio file

language · string

The detected or specified language code in ISO-639-1 format (e.g., 'en', 'es', 'fr')

confidence · number

Confidence score from 0 to 1 where 1 indicates highest confidence in transcription accuracy

speakers · object[]

Speaker diarization segments, returned only when include_speaker_data is true.

Show child attributes