API Status: Operational

Dhivehi speech-to-text, refined.

Fast, production-grade transcription API optimized specifically for the Dhivehi language. Powered by fine-tuned models for outstanding precision and structural accuracy.

dhivehgpt-console.log
Response Time
~180ms
Model Accuracy
98.4%
[11:45:01] INFO System initialized. Listening for jobs...
[11:45:12] JOBS New job '8b8f2d...' submitted successfully.
[11:45:13] BEAM Splitting audio into chunks... (4 chunks found)
[11:45:32] COMP Job completed. 100% transcript computed.
🔑

Bearer Authentication Required

All API endpoints are strictly secured. You must pass your api key in the headers as a Bearer token: Authorization: Bearer <YOUR_API_TOKEN>

Endpoints Reference

Easily integrate DhivehGPT transcription into any code backend with two core API operations.

POST /api/transcriptions

Submit a new audio URL to be transcribed. This operation is asynchronous. The system will start chunking the file and transcribing chunks in parallel, then invoke your callback webhook when complete.

Request Parameters

Name Type Required Description
audio_url string required Fully-qualified URL to the raw audio/video file.
callback_url string required Fully-qualified HTTPS webhook endpoint that receives notifications.
# Submit a transcription job via cURL
curl -X POST https://api.dhivehgpt.mv/api/transcriptions \
  -H "Authorization: Bearer YOUR_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "audio_url": "https://example.com/audio.mp3",
    "callback_url": "https://yourdomain.com/webhook"
  }'
GET /api/transcriptions/{uuid}

Retrieve the state and progress of an active transcription job. Once the transcription has successfully completed, the response payload will include the fully compiled Dhivehi transcript.

Path Parameters

Name Type Required Description
uuid string required The transcription job's unique identifier.
# Check job status via cURL
curl -X GET https://api.dhivehgpt.mv/api/transcriptions/JOB_UUID \
  -H "Authorization: Bearer YOUR_API_TOKEN"