Text & Extraction API

POST /chat/completions — an OpenAI-compatible text endpoint, built as the side channel for voice agents: everything that reasons about a conversation instead of speaking in it. Structured extraction, outcome classification, post-call QA scoring, voicemail detection, idle re-prompts, chat-mode testing — one endpoint powers them all, with native Hindi, Hinglish and code-switched comprehension.

Point any OpenAI SDK at base URL https://platform.oogam.ai/v1 with your existing sk-setu-… key. A Naad (voice) key can call naad-text-v1 directly — no extra product signup.

Structured extraction

Set response_format to {"type": "json_object"} and the reply's content is guaranteed parseable JSON — the platform validates it server-side (and retries the model once if needed) before answering, so JSON.parse on the content never throws. Combined with temperature: 0 the output is deterministic: the same transcript yields the same record.

curl https://platform.oogam.ai/v1/chat/completions \
  -H "Authorization: Bearer $OOGAM_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "naad-text-v1",
    "temperature": 0,
    "response_format": {"type": "json_object"},
    "messages": [
      {"role": "system",
       "content": "Return JSON with keys date (YYYY-MM-DD) and time (HH:MM, 24h)."},
      {"role": "user",
       "content": "25 अगस्त 2026 को शाम 5 बजे का appointment book कर दो"}
    ]
  }'

Tip: state dates explicitly (or pass today's date in your system prompt and post-process relative words like कल/परसों yourself) — explicit anchors extract deterministically.

Response — standard shape, with real token usage for cost attribution:

{
  "id": "chatcmpl-1e97a0ee22d04527b805315c",
  "object": "chat.completion",
  "created": 1755763200,
  "model": "naad-text-v1",
  "choices": [{
    "index": 0,
    "message": {
      "role": "assistant",
      "content": "{\"date\": \"2026-08-25\", \"time\": \"17:00\"}"
    },
    "finish_reason": "stop"
  }],
  "usage": {"prompt_tokens": 81, "completion_tokens": 29, "total_tokens": 110}
}

Request parameters

Response (non-streaming): the object shown above — id, choices[0].message.content, and a usage block with real token counts. With json_object, content is a JSON object string; parse it directly.

Streaming

With stream: true the response is text/event-stream: a first chunk carrying delta.role, then content chunks, a final chunk with finish_reason: "stop", and the literal data: [DONE] sentinel. Each event is a chat.completion.chunk. Note the stream does not include a usage block — read token usage from a non-streaming call, or from your dashboard. Most voice-agent tasks (extraction, classification) are batch, so streaming is optional.

data: {"id":"chatcmpl-…","object":"chat.completion.chunk","choices":[{"index":0,"delta":{"role":"assistant"},"finish_reason":null}]}

data: {"id":"chatcmpl-…","object":"chat.completion.chunk","choices":[{"index":0,"delta":{"content":"नम"},"finish_reason":null}]}

data: {"id":"chatcmpl-…","object":"chat.completion.chunk","choices":[{"index":0,"delta":{"content":"स्ते"},"finish_reason":null}]}

data: {"id":"chatcmpl-…","object":"chat.completion.chunk","choices":[{"index":0,"delta":{},"finish_reason":"stop"}]}

data: [DONE]

Endpoint errors

Everything else (auth, rate limits, wallet) uses the shared error contract.

Embeddings

POST /embeddings — OpenAI-compatible, model naad-embed-v1, for building retrieval over Indian-language content without shipping documents offshore.

curl https://platform.oogam.ai/v1/embeddings \
  -H "Authorization: Bearer $OOGAM_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "naad-embed-v1",
    "input": ["ऑर्डर कहाँ है?", "मुझे रिफंड चाहिए"]
  }'

Response: data holds one embedding per input, in order, each with its index. The vector dimension is fixed per model version and published at activation — it is a stability contract, so declare it once in your vector(N) column.

{
  "object": "list",
  "data": [
    { "object": "embedding", "index": 0, "embedding": [0.013, -0.221, …] },
    { "object": "embedding", "index": 1, "embedding": [-0.047, 0.164, …] }
  ],
  "model": "naad-embed-v1",
  "usage": { "prompt_tokens": 12, "total_tokens": 12 }
}

Status: activating soon. Until an embeddings provider is enabled on the deployment, the endpoint answers 503 model_not_configured (JSON, with a Retry-After header) — never a mock vector, which would silently corrupt your index. Over-limit batches return 400 batch_too_large; over-long inputs 400 input_too_long; empty input 400 missing_input.

Latency & determinism

• Typical extraction over a few-hundred-token transcript: ~1 second.
• temperature: 0 + same input → same output, run after run.
• usage reports the model's own token counts — bill and attribute per call.