Quickstart

Make your first request in two minutes with any OpenAI or Anthropic SDK.

このページはまだ翻訳されていません。英語版を表示しています。

1. Get a key

Create an API key on the API keys page. Keys start with xpl_ and are shown once. Paste it into Add API key at the top of this page and every example fills it in.

export EXPLABS_API_KEY="YOUR_API_KEY"

2. Send a request

Point any OpenAI SDK at the base URL. Nothing else changes. Model ids are slugs from GET /v1/models.

POST/v1/chat/completions
curl https://api.experientiallabs.ai/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "gpt-6-luna", "messages": [{"role": "user", "content": "Hello"}]}'

3. Or use the Anthropic SDK

The Anthropic SDKs take the bare origin (no /v1); they append /v1/messages themselves. Any catalog slug works, not just Claude.

POST/v1/messages
from anthropic import Anthropic

client = Anthropic(base_url="https://api.experientiallabs.ai", api_key="YOUR_API_KEY")
msg = client.messages.create(model="gpt-6-luna", max_tokens=1024, messages=[{"role": "user", "content": "Hello"}])
print(msg.content[0].text)

What comes back

A standard response plus two gateway fields: usage.cost (USD charged) and top-level provider (the provider that served it). The x-request-id header identifies the call in your usage. Next: Coding agents, Errors, How routing works.

{
  "id": "chatcmpl-...",
  "model": "gpt-6-luna",
  "provider": "openai",
  "choices": [{"message": {"role": "assistant", "content": "Hello!"}, "finish_reason": "stop"}],
  "usage": {"prompt_tokens": 8, "completion_tokens": 3, "total_tokens": 11, "cost": 0.000041, "is_byok": false}
}