Glossary 

AI provider

External service that processes prompts and returns responses (for example OpenAI, Gemini, Anthropic, Mistral).

Model

Specific AI model identifier within a provider.

Embedding

Numeric vector representation of text used for similarity and semantic tasks.

Request orchestration

The process of building payloads, sending requests, and parsing responses.

Basic Auth

HTTP authentication using username and password, encoded as Authorization: Basic ....

Cache

Local storage used to reduce repeated API calls and improve response times.

Service layer

Internal reusable classes used by other extensions rather than direct frontend output.