About Dukou API

One API key connects your application to leading models from DeepSeek, Claude, GPT and Gemini. Use OpenAI- or Anthropic-compatible APIs, pay for what you use, and check your balance at any time.

What we do

Dukou API is a unified gateway for AI models. Point your existing client at our endpoint and choose a model; in most cases, your integration needs no other changes.

Why we built it

We built Dukou API to reduce the friction of trying multiple AI providers: account and payment requirements, balances scattered across services, incompatible SDK conventions, and bills that are hard to reconcile. Spend your time building; manage model access and usage in one place.

How it works

  1. Authenticate: We validate your API key and available balance.
  2. Route: We send the request to the upstream provider for the model you selected.
  3. Record usage: We calculate the charge from token usage and the model's listed rate, then record it against your balance.

Responses follow the compatible API format, including server-sent events (SSE) for streaming. Model names are Dukou API identifiers; see the Model Catalog for current listings. We are an independent third-party gateway, not an official provider channel.

Who it's for

Pricing

Charges are based on token usage. Input, output and supported cache usage may have different rates. The Model Catalog publishes current rates, including group multipliers. There are no hidden fees, monthly subscriptions or minimum spend. New accounts receive trial credits.

Data and security

Read our Terms of Service and Privacy Policy.

Contact

For integration questions, billing enquiries or incident reports, contact us through the support ticket area in your console. Include the original error message when reporting a problem.

Get started

Sign up to receive trial credits. No payment card or approval wait is required.

Open the console · Browse models and rates