RAG
Memory
Memory Lab
Gold Lab
Compare
Playground
Eval
Side-by-side comparison
Same query through 2–4 configs · diff quality × cost × latency
Query
Почему я снова застрял с Shadow и этой теорией?
Route mode
auto
rag
memory
long_context
hybrid
topK: 3
maxContextTokens
Route mode
auto
rag
memory
long_context
hybrid
topK: 10
maxContextTokens
Add config
Run comparison