Get started / QuickstartSign inGet an API key

Quickstart

Create a key, send a recording, read the transcript.

1. Get a key

Sign in to the console, open your project and choose API keys → Create key. The secret is shown once; keep it in an environment variable. The same page shows your project's API address — every request goes there.

Shell
export TWENTYSIX_API_KEY="sk_…"
export TWENTYSIX_BASE_URL="https://…"   # your project's API address

2. Transcribe a short clip

Audio up to a few minutes can be sent in one request and the transcript comes back in the response.

curl "$TWENTYSIX_BASE_URL/v1/audio/transcriptions" \
  -H "Authorization: Bearer $TWENTYSIX_API_KEY" \
  -H "Idempotency-Key: $(uuidgen)" \
  -F file=@clip.wav \
  -F language=hi-IN \
  -F mode=codemix

3. Transcribe a long recording

For anything longer — up to two hours — upload the file, start a transcription and wait for it. The SDKs do all of it in one call.

result = client.transcribe_file("meeting.m4a", "ta-IN", diarization=True)
for s in result["segments"]:
    print(f'{s["start"]:7.1f}  {s.get("speaker", "")}  {s["text"]}')

The step-by-step version, with plain HTTP, is in Long recordings.

What you get back

JSON
{
  "id": "tr_…",
  "object": "transcription.result",
  "status": "completed",
  "mode": "codemix",
  "duration_seconds": 41.6,
  "channels": 1,
  "languages": ["hi-IN"],
  "text": "कल की meeting का time बदलकर 4 बजे कर दीजिए…",
  "segments": [
    { "id": 0, "start": 0.0, "end": 6.2, "text": "कल की meeting का time बदलकर 4 बजे कर दीजिए…" }
  ],
  "warnings": []
}

Every field is described in Results and timestamps.

Next

Scriba is a product of 26AI Private Limited, Pune, India.© 2026 26AI Private Limited