DupliChat API

The DupliChat API is OpenAI-compatible. Editors and agents that accept a custom OpenAI base URL (Cursor, Cline, Continue, Roo Code and others) can use your plan's credits with GPT models. It is included in paid plans (Starter, Plus and Pro); Free and trial accounts cannot create keys.

Quick start

Log in and open Plans & billing → DupliChat API, then create a key. It starts with dbk- and is shown once.
Set the base URL to https://duplichat.com/v1 and use the key as the API key.
Pick a model your plan allows: GET https://duplichat.com/v1/models lists them.

curl https://duplichat.com/v1/chat/completions \
  -H "Authorization: Bearer dbk-YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-5.6-luna","messages":[{"role":"user","content":"Hello"}]}'

Endpoints

GET /v1/models: the models your key may use.
POST /v1/chat/completions: chat completions with stream, tools and tool_choice, image input, temperature, max_tokens and the other common parameters. One answer per request (n = 1), requests up to 4 MB.

Not available: embeddings, image generation, audio and the Responses API.

Set up your tool

Cursor: Settings → Models → OpenAI API Key: paste the key, switch on "Override OpenAI Base URL" and enter https://duplichat.com/v1. Add the model names you want.
Cline, Continue, Roo Code (VS Code): choose the provider "OpenAI Compatible", enter the base URL and key, then a model name.
Other tools: anything that lets you set a custom OpenAI-compatible base URL and key works.

Credits, limits and errors

API calls use the same credits as chat, by tokens and model. See pricing.
401: key missing, wrong or revoked. 403: plan without API access, or a model above your plan. 429 with code insufficient_quota: credits used up. 429 otherwise: slow down.
Keep keys secret: anyone with a key spends your credits. Revoke a key in Plans & billing at any time.

Questions? See the FAQ or the contact page.