← BACK TO AI LAB

AI LAB / RAG / Reranking

Reranker Latency Budget

Measuring how many candidates a reranker can score before response latency stops being worthwhile.

RAGRerankingPerformance
EXPERIMENT
EXP-020
STATUS
complete
DATE
Jul 22, 2026
TAGS
RAG · Reranking · Performance

EXPERIMENT / HYPOTHESIS

Reranking 20 candidates gives most of the quality gain within a 250ms interactive latency budget.

  1. 01Retrieve 10, 20, 40, and 80 candidates
  2. 02Measure NDCG and end-to-end latency
  3. 03Repeat across short and long documents
Best candidate count
20
Added P95
184ms
NDCG gain
+11.2%

CONCLUSIONTwenty candidates captured most quality improvement. Larger sets produced diminishing gains and unstable tail latency.

01

Budget

The experiment assigned 250ms of an 800ms retrieval budget to reranking, including serialization and network overhead.

  • Warm and cold runs
  • Batch sizes
  • Tail latency
02

Decision

The production default is 20 candidates, with lower counts for simple queries and asynchronous reranking for research mode.

  • Adaptive count
  • Budget-aware fallback
  • Observe P95