Your Users Speak.
The Stars Answer.

# Plain-text question (no audio)? Use the text query endpoint instead.
# /api/voice/binary takes raw AUDIO only — see the cURL (audio) tab.
curl -X POST https://api.vedika.io/api/v1/astrology/query \
  -H "Authorization: Bearer vk_live_xxx" \
  -H "Content-Type: application/json" \
  -d '{
  "question": "What does my chart say about career this year?",
  "birthDetails": {"datetime": "1990-04-12T08:30:00", "latitude": 28.61, "longitude": 77.21, "timezone": "+05:30"}
}'

# Returns JSON text. Pipe that text to /api/voice/binary for spoken audio.

# Success response (audio/mpeg binary)
HTTP/1.1 200 OK
Content-Type: audio/mpeg
X-Vedika-Voice-Duration: 12.4
X-Vedika-Voice-Cost: 0.0213

<binary audio bytes — pipe to player or save to file>

# Error responses
# Empty / unreadable audio body:
HTTP/1.1 400 Bad Request
{ "success": false, "code": "NO_AUDIO", "error": "Empty audio body..." }

# Wallet can't cover the call:
HTTP/1.1 402 Payment Required

# Voice requires a Business or Enterprise plan:
HTTP/1.1 403 Forbidden   # code: VOICE_PLAN_REQUIRED

# An unknown x-vedika-voice-tier does NOT error — it falls back to Standard.

# Crisis detection → deterministic helpline JSON, $0.00, no model call
HTTP/1.1 200 OK
Content-Type: application/json

{
  "answer": "...iCall (India): 9152987821 · Vandrevala: 1860-2662-345...",
  "cost": { "costUsd": 0 },
  "safety": "crisis-helpline",
  "success": false
}