Quickstart

OpenDunes is an OpenAI-compatible API. If you've used any OpenAI SDK before, you already know how to use it — just swap the base URL and key.

Get an API key#

Sign in and go to Dashboard → API Keys. A default key is created automatically on signup. You can create additional keys, set per-key rate limits, and revoke them at any time.

Copy your key and store it as an environment variable — never hardcode it in source:

export OPENDUNES_API_KEY="sk-0a1b2c3d4e5f..."

Make your first request#

Pick any model from the catalog and call POST /v1/chat/completions. Here's the simplest possible request using three different clients:

curl https://opendunes.com/api/v1/chat/completions \
  -H "Authorization: Bearer $OPENDUNES_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "anthropic/claude-sonnet-5",
    "messages": [{ "role": "user", "content": "Marhba! Say hello in one line." }]
  }'

Use the OpenAI SDK#

Any OpenAI SDK works out of the box. Set base_url (Python) or baseURL (Node/TS) to https://opendunes.com/api/v1 and pass your OpenDunes key as the API key. No other changes needed.

The SDK's full request surface is honored by OpenDunes today — model, messages, stream, max_tokens, temperature, top_p, tools, tool_choice, response_format, stop, seed, penalties, logprobs, and more, plus OpenDunes extensions like models fallbacks and provider routing. See Chat Completions for the full parameter reference.

Stream the response#

Add "stream": true and the server will push tokens over server-sent events as they arrive. See Streaming for the full SSE protocol and how to handle mid-stream errors.

Next steps#

• Models — browse the full catalog, filter by capability, check DA pricing
• Chat Completions — every supported request parameter
• Credits & Pricing — how DA billing and micro-DA work