Skip to main content

Onboarding (LLM API)

Welcome! This guide explains the basic concepts, prerequisites, and steps to get started with the CAIP 2.0 LLM API.

1 - Create CAIP Space​

First, you need a CAIP Space, the logical container for all CAIP Feautures. You can find your CAIP Spaces in our Self-Service Portal or directly at https://caip.bmw.cloud.

Please raise a Service Request here to create a new CAIP Space, if you do not have a CAIP Space created for your use case yet.

2 - Create API Key​

To obtain an API Key for your CAIP Space, you can use the Self-Service Portal or raise a Service Request of type 'Request API Key'. Our team will generate and provide you with a UUID-format API key that is associated with your CAIP Space. This key can be used to access CAIP APIs, such as the LLM API, and to integrate with external tools or programmatic workflows securely.

UUID Format Required (RoW)

Only UUID-format keys (e.g. 123e4567-e89b-12d3-a456-426614174000) are accepted by the RoW LLM API. Old-format keys are deprecated. If your key does not match this format, please request a new one.

3 - Consume LLM API​

Once you have received your API key, you can use it to access the CAIP LLM API for chat, embeddings, and image generation use cases. The API key must be included in the Authorization header of each request. For best compatibility, use the Bearer prefix:

Authorization: ******

Regional Endpoints​

There are two separate LLM API instances:

  • EMEA / Rest of World (RoW): https://llm.api.caip.bmw.cloud
  • China: https://llm.api.caip.bmwchina.cloud

This separation is required due to local regulatory requirements in China. Please use the correct endpoint for your region.

How to Use​

  • Refer to the LLM API Documentation for detailed usage instructions, supported models, and integration examples (including Python SDK and curl).
  • For a full list of endpoints, request/response schemas, and error codes, see the LLM API Reference.

Example: Chat Completion Request (OpenAI Compatible Format)

curl -X POST \
  'https://llm.api.caip.bmw.cloud/v1/chat/completions' \
  -H 'Authorization: ******' \
  -H 'Content-Type: application/json' \
  -d '{
    "model": "gpt-4o",
    "messages": [
      {"role": "user", "content": "Hello!"}
    ]
  }'

For more code samples and advanced usage (including streaming, embeddings, and image generation), consult the documentation and API reference linked above.