[ Docs ]

Quick Start

Prototype in Quark, then shape Origin and NextGen requests against the API contract.

Start with Quark

The fastest way to use IdonAI today is Quark — the live assistant powered by the same Origin and NextGen models you will call over the API.

Use Quark to:

Validate prompts and system instructions before you wire an integration
Compare origin vs nextgen on the same task
Confirm latency and output style for your use case

API keys are under development. You can prepare integrations against the contract below; public key issuance is not available yet.

Request shape

When keys ship, chat completions will use a familiar REST shape against https://api.idonai.com.

typescript

curl https://api.idonai.com/v1/chat/completions \
  -H "Authorization: Bearer $IDONAI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "origin",
    "messages": [
      {"role": "user", "content": "Explain quantum entanglement."}
    ],
    "max_tokens": 1024
  }'

Model ids

Id	Role
`origin`	Fast, production-default model for chat and tools
`nextgen`	Higher-capability model for harder reasoning and longer work

Pass the id as the model field. The same ids appear in Quark and in the API reference.

Streaming

Set "stream": true on the chat completions body to receive SSE chunks. See the API reference for chunk shape and termination (data: [DONE]).

Auth and limits

Authentication — Bearer tokens, env vars, and what not to put in the browser
Rate limits — Developer / Pro / Enterprise tiers aligned with pricing

SDKs

Official clients for Python, TypeScript, Go, and Java are planned. Package names are reserved; public installs land when API keys ship.

SDKs overview
Python · TypeScript · Go · Java

Until then, prefer Quark for product exploration and REST sketches for integration design.

Next steps

Prototype in Quark
Read Authentication and Rate limits
Bookmark the API reference and SDKs