Integrations

Use Aether as the retrieval layer for your favorite AI provider.

Aether fits naturally into the RAG (Retrieval-Augmented Generation) pattern: ingest your documents into the object store, retrieve relevant context at query time, then pass that context to an LLM to get grounded, citation-ready answers.

Without RAG, an LLM can only answer based on what it learned during training — which might be outdated or missing your specific data. With Aether in the loop, the LLM gets fresh, relevant context from your own documents every time a question is asked. This dramatically reduces hallucinations and keeps answers grounded in facts.

The key method is retrieve(). It combines vector search with content download in a single call, returning the chunks you need to build a prompt:

results = client.retrieve("How does the billing system work?", top_k=5)

From there, you pass results into any LLM provider as context. The guides below show exactly how to wire this up for each provider.

Building with an AI app builder?

If you're wiring Aether in from a coding agent or app builder, start with the per-platform on-ramp — a one-paragraph intro, a ready-to-paste prompt, a manual fallback, and example builds, all tuned to that platform.