Embedding Router & Universal Vector Space
Switch embedding models instantly. Migrate without re-embedding. OpenAI, Gemini — one unified API.
Every embedding model creates a different vector space. Switching providers means re-embedding your entire corpus — millions of vectors, days of compute, production downtime.
Schift solves this. Our projection matrices transform existing vectors into any target model's space with 99.7% retrieval recovery, in minutes instead of days.
from schift import Schift
s = Schift(api_key="sk-...")
# Embed with any provider through one API
vec = s.embed("quarterly revenue report", model="openai/text-embedding-3-large")
# Switch your entire DB to a new model — no re-embedding
s.upgrade(db="my_pgvector", to="google/gemini-embedding-2") # 99.7% recovery