Build with AI
Let your AI assistant write the integration. Copy the prompt below — it bundles the complete API reference plus instructions — and paste it into ChatGPT, Claude or Cursor.
Copies a prompt + the full reference from /llms-full.txt
How it works#
The copied prompt tells the assistant to use only our reference (no invented endpoints), to prefer the one-call synchronous endpoint POST /v1/generate?wait=true, and to read your key from API18_KEY. Then just ask it for what you want — e.g. “generate a 5s video from this image”.
Tip: paste the prompt first, then describe your use case in plain language. The assistant already has every model id, parameter and example it needs.
Per-model files#
Only using one model? Each model page has its own llms.txt (smaller, focused) — open any model and click “Copy llms.txt”. The index lives at https://api18.dev/llms.txt.