Voyage AI Rerank 2.5 Lite
Latency-optimized version maintaining instruction-following and 32K context capabilities with streamlined inference. Designed for high-volume production deployments prioritizing cost efficiency over maximum accuracy.
Model Information
- Provider
- Voyage AI
- License
- Proprietary
- Price per 1M tokens
- $0.020
- Release Date
- 2025-08-11
- Model Name
- voyage-rerank-2.5-lite
- Total Evaluations
- 2100
Performance Record
Wins1058 (50.4%)
Losses961 (45.8%)
Ties81 (3.9%)
Wins
Losses
Ties
Performance Overview
ELO ratings by dataset
Voyage AI Rerank 2.5 Lite's ELO performance varies across different benchmark datasets, showing its strengths in specific domains.
Voyage AI Rerank 2.5 Lite - ELO by Dataset
Detailed Metrics
Dataset breakdown
Performance metrics across different benchmark datasets, including accuracy and latency percentiles.
PG
ELO 160652.3% WR183W-167L-0T
Latency Distribution
- Mean
- 637ms
- P50 (Median)
- 614ms
- P90
- 817ms
business reports
ELO 158957.1% WR200W-141L-9T
Latency Distribution
- Mean
- 580ms
- P50 (Median)
- 607ms
- P90
- 816ms
FiQa
ELO 156552.3% WR183W-159L-8T
Accuracy Metrics
- nDCG@5
- 0.111
- nDCG@10
- 0.122
- Recall@5
- 0.103
- Recall@10
- 0.135
Latency Distribution
- Mean
- 686ms
- P50 (Median)
- 611ms
- P90
- 829ms
MSMARCO
ELO 152747.1% WR165W-124L-61T
Accuracy Metrics
- nDCG@5
- 0.981
- nDCG@10
- 0.983
- Recall@5
- 0.993
- Recall@10
- 1.000
Latency Distribution
- Mean
- 542ms
- P50 (Median)
- 611ms
- P90
- 635ms
DBPedia
ELO 150257.1% WR200W-147L-3T
Accuracy Metrics
- nDCG@5
- 0.692
- nDCG@10
- 0.763
- Recall@5
- 0.064
- Recall@10
- 0.111
Latency Distribution
- Mean
- 555ms
- P50 (Median)
- 605ms
- P90
- 664ms
Scifact
ELO 140536.3% WR127W-223L-0T
Accuracy Metrics
- nDCG@5
- 0.844
- nDCG@10
- 0.849
- Recall@5
- 0.892
- Recall@10
- 0.900
Latency Distribution
- Mean
- 642ms
- P50 (Median)
- 618ms
- P90
- 816ms
Compare Models
See how it stacks up
Compare Voyage AI Rerank 2.5 Lite with other top rerankers to understand the differences in performance, accuracy, and latency.