Speech-to-Text | OpenDunes Docs

Speech-to-Text (STT) converts audio into transcribed text. OpenDunes supports Arabic transcription natively — including Algerian Darija — alongside French and English.

OpenDunes takes the audio inline as base64 JSON (not a multipart file upload), so call the endpoint directly with httpx rather than the OpenAI SDK's audio.transcriptions helper.

Transcribe an audio file#

Python

import os
import base64
import httpx

headers = {"Authorization": f"Bearer {os.environ['OPENDUNES_API_KEY']}"}

with open("recording.mp3", "rb") as f:
    data = base64.b64encode(f.read()).decode()

resp = httpx.post(
    "https://opendunes.com/v1/audio/transcriptions",
    headers=headers,
    json={
        "model": "example-provider/stt-model",
        "input_audio": {"data": data, "format": "mp3"},
        "language": "ar",   # optional — improves accuracy
    },
)
print(resp.json()["text"])

Parameters#

Parameter	Description
`model`	STT model slug (`?modality=transcription`).
`input_audio.data`	Base64-encoded audio bytes.
`input_audio.format`	One of `wav`, `mp3`, `flac`, `m4a`, `ogg`, `webm`, `aac`.
`language`	BCP-47 language hint, e.g. `"ar"`, `"fr"`, `"en"`. Optional.
`temperature`	0–1. Lower values produce more deterministic output. Optional.

Transcription is billed per second of audio in DA, rounded up. See the STT API reference.

Speech-to-Text (STT) converts audio into transcribed text. OpenDunes supports Arabic transcription natively — including Algerian Darija — alongside French and English.

OpenDunes takes the audio inline as base64 JSON (not a multipart file upload), so call the endpoint directly with httpx rather than the OpenAI SDK's audio.transcriptions helper.

Transcribe an audio file#

Python

import os
import base64
import httpx

headers = {"Authorization": f"Bearer {os.environ['OPENDUNES_API_KEY']}"}

with open("recording.mp3", "rb") as f:
    data = base64.b64encode(f.read()).decode()

resp = httpx.post(
    "https://opendunes.com/v1/audio/transcriptions",
    headers=headers,
    json={
        "model": "example-provider/stt-model",
        "input_audio": {"data": data, "format": "mp3"},
        "language": "ar",   # optional — improves accuracy
    },
)
print(resp.json()["text"])

Parameters#

Parameter	Description
`model`	STT model slug (`?modality=transcription`).
`input_audio.data`	Base64-encoded audio bytes.
`input_audio.format`	One of `wav`, `mp3`, `flac`, `m4a`, `ogg`, `webm`, `aac`.
`language`	BCP-47 language hint, e.g. `"ar"`, `"fr"`, `"en"`. Optional.
`temperature`	0–1. Lower values produce more deterministic output. Optional.

Transcription is billed per second of audio in DA, rounded up. See the STT API reference.