🤗 Inference Endpoints can be used through the UI and programmatically through an API. The API exposes open-API specification for each available route.
<iframe src="https://api.endpoints.huggingface.cloud/" style='height: 60vh; width: 100%;' frameborder="0" id="iframe" frameBorder="0">Browser not compatible.</iframe>Update May 2024: We have renamed instances and further details can be found in the pricing documentation. For example, when using 1x A10G instance, the naming is:
- instance_type: nvidia-a10g
- instance_size: x1