Region Inference

The base URL you call decides where your request is served. Nothing in your account, your key or the request body selects a region. The default GreenPT endpoint always uses EU inference; to use US inference, call the US endpoint explicitly.

Regional endpoints

Your prompts and outputs are processed in the region of the endpoint you call. Account, billing and usage data stay in the EU whichever endpoint you use. See Data Processors for the providers behind each region.

Set the base URL in your OpenAI-compatible client, or use the regional hostname in HTTP requests. Keep the /v1 prefix when configuring a client. One API key works on every endpoint.

Model availability

The EU endpoints serve the full model catalog. The US endpoint serves these models, under the same ids the EU endpoint uses:

On the US endpoint, qwen3-embedding-8b accepts up to 8,176 input tokens.

Prices are per endpoint: a model served from the US costs what that route costs, which is not the price the EU endpoint charges for the same id. See the US tab on Pricing, or call GET /v1/pricing on the base URL you use, which is the authoritative price list for that endpoint.

Use the EU / US filter in Model Cards to compare availability, capabilities, and pricing. Region and category filters work together to show models available in the selected region.

US endpoint scope

These are the endpoints the US host serves:

Every other endpoint answers 403 there with {"code": "endpoint_not_available_in_location"}, and names the host to call instead. Asking the US endpoint for a model it does not serve answers 403 with {"code": "no_offering_in_allowed_locations"} and the locations that model is offered in, so a missing id has one call that says why.

Call a US model

With your API key stored in GREENPT_API_KEY, send a chat-completion request to the US endpoint:

curl https://api.us.greenpt.ai/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $GREENPT_API_KEY" \
  -d '{
    "model": "deepseek-v4.1-flash",
    "messages": [
      { "role": "user", "content": "Explain how a solar panel works." }
    ]
  }'

For US embeddings, select qwen3-embedding-8b:

curl https://api.us.greenpt.ai/v1/embeddings \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $GREENPT_API_KEY" \
  -d '{
    "model": "qwen3-embedding-8b",
    "input": "Renewable energy powers sustainable AI."
  }'

List regional models

Query the Models API on the endpoint you intend to use:

curl https://api.us.greenpt.ai/v1/models \
  -H "Authorization: Bearer $GREENPT_API_KEY"

For EU inference, use https://api.greenpt.ai/v1/models.

For processing-location details, see Data Processing Locations.