Documentation
Loading…
Completions and embeddings
The legacy text completion endpoint and embeddings.
Base URL for this deployment: https://www.aiteraiworld.com/api/v1
POST /completions
Accepts prompt instead of messages and returns text_completion choices, including as a stream. Routing, billing and headers are identical to chat.
json
{ "model": "author/model", "prompt": "Once upon a time", "max_tokens": 64 }POST /embeddings
Takes a string, a list of strings or token arrays as input, with encoding_format of float or base64. Input tokens are metered at the model's prompt rate even though there is no completion.
json
{ "model": "author/embedding-model", "input": ["first document", "second document"] }