OpenAI-Compatible REST API

Integrate Not-Live Ai directly into your applications, Python scripts, or LangChain pipelines using drop-in endpoints. Authenticated developers receive free, unlimited named API keys.

Sign In to Generate Custom API Keys

Only logged-in developers can create, manage, and revoke named API keys. Sign in or create a free account to get your keys instantly.

Request Body Reference (JSON)

Parameter Type Default Description
model string not-live-balanced Options: not-live-fast (Fast: direct inference, no online search), not-live-balanced (Balanced: adaptive reasoning with live web search), not-live-deep (Deep Research: multi-source live search with exhaustive CoT)
messages array required Array of message objects with role (system/user/assistant) and content
stream boolean true When true, responses stream via Server-Sent Events (SSE) token-by-token
enableThinking boolean true Includes internal multi-step verification and reasoning trace tokens
temperature float 0.6 Sampling temperature between 0.0 and 1.0 (lower values are more deterministic)
max_tokens integer 4096 Maximum number of tokens to generate per completion (up to 16,384)