What akasha provides

akasha exposes a small set of public entry points for common model workflows:

Capability	Entry point	Use it when...
Chat and question answering	akasha.ask()	You need an answer from a model, optionally with context.
Retrieval-augmented generation	akasha.RAG()	You need to load and search documents before answering.
Agents	akasha.agents()	You need tool calling or a multi-step task.
Summaries	akasha.summary()	You need to summarize text, files, or URLs.
Long-term memory	MemoryManager	You need to store and retrieve semantic memories.

The public API hides provider-specific construction behind aliases such as gemini:, openai:, anthropic:, and ollama:.

A useful mental model

Question → optional context → model → answer
                         ↘ tools / retrieval / memory

Read Choosing an API when you are unsure which entry point to use.