India's cheapest Indic ASR,
Hindi & 20+ languages at ₹10/hr.
Enterprise-grade Hindi and Indian language speech-to-text at just ₹10 per audio-hour (~16 paise/min). Speaker diarization, Indic PII redaction, and segment-level timestamps in one API call — up to 78% cheaper than existing commercial APIs.
- S1·00:12नमस्ते, आपकी क्या समस्या है?
- S2·00:15हैलो, मेरा order अब तक नहीं आया।
- S1·00:23एक minute, मैं देखता हूँ। आपका numberPHONE बताइए —
- S2·00:31Yes, and please also mail me the invoice.
- S3·00:37Supervisor here — how can I help?
Every knob you need, one POST.
Multipart audio + a JSON body. Back comes a diarized, PII-safe transcript with segment-level timestamps. Everything the enterprise batch APIs charge extra for — at a fraction of the price.
- Familiar HTTP multipart transport — no SDK required
- Idempotency keys, webhook auth headers, multi-channel stereo
- Batch SLA: results within 30 minutes (P50 ~15–20 min)
# Upload audio, get back a diarized, PII-safe transcript.
curl -X POST https://api.asrcore.com/v1/jobs \
-H "Authorization: Bearer $ASR_INSIGHTS_KEY" \
-F "[email protected]" \
-F params={
"language_codes": ["hi", "en"],
"diarize": true,
"min_speakers": 2,
"pii": {"detect": true, "redact": "text"},
"output_format": "json"
}
Up to ~10× cheaper per audio-hour.
Same batch shape as every enterprise ASR API. A fraction of the unit economics because we built the pipeline lean from day one.
Pricing revealed at launch. Get early access →
Turn on only what you need.
Base transcription is one price. Diarization, Indic PII, custom vocabulary, and delivery add-ons are priced separately and enabled per request — pay only for what a job actually uses.
Who spoke when. Configurable min/max speakers. Accurate segment-level timestamps per speaker turn.
Detect and redact PAN, Aadhaar, phone, UPI, GSTIN in Hindi, English, and mixed-script audio.
Transcribe agent and customer stereo channels independently with pristine zero-bleed channel separation.
Pass language_codes: ["hi", "en"] for Hinglish and code-mixed calls without splitting audio.
SRT, VTT, JSON with timestamps, or plain text. One request, one flag.
Async job completion delivered to your endpoint with retry logic and custom auth headers.
From BPO QA to compliance audits.
One batch endpoint, six workloads teams already run on it — contact-center QA, sales call review, banking compliance, legal transcripts, media captions, and multi-language research.
Start transcribing in Hindi + 18 more languages
Batch tier. Launching soon — get early access.