Library / SDK
Guardrails & StructureRetrieval & Memory
Semantic Router
1Fast 2semantic 3decision-making 4layer 5for 6LLMs
the six Ws · specification
W1
Who
Maintained by Aurelio AI (aurelio-labs).
W2
What
A library using embeddings to make fast semantic routing decisions instead of slow LLM calls.
W3
Where
A Python library embedded in your application.
W4
When
Reach for it when you need low-latency intent routing or tool selection.
W5
Why
Replaces slow, costly LLM-based routing with deterministic embedding similarity.
W6
With
Works with OpenAI, Cohere, and Hugging Face embeddings; optional vector stores.
W7
Watch
Maintained by Aurelio AI; open-source (MIT) library, self-hostable. It only calls whichever embedding provider you configure.
alternatives
works with