# 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. If you want to compare the best rerankers for your data, try [Agentset](https://app.agentset.ai/).

## Leaderboard Rank
#6 of 12

### ELO Rating

1520

#6

### Win Rate

53.4%

#6

### Accuracy (nDCG@10)

0.103

#4

### Latency

616ms

#8

### 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**: 3300

### Performance Record
- **Wins**: 1763 (53.4%)  
- **Losses**: 1416 (42.9%)  
- **Ties**: 121 (3.7%)

## Rerankers Are Just One Piece of RAG
Agentset gives you a managed RAG pipeline with the top-ranked models and best practices baked in. No infrastructure to maintain, no reranking to configure.

### 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
14291479152915791614arguanabusiness reportsFiQaMSMARCODBPediaPG

## Detailed Metrics

### Dataset breakdown
Performance metrics across different benchmark datasets, including accuracy and latency percentiles.

#### arguana
- **ELO**: 161463.6%  
- **WR**: 350W-199L-1T

##### Accuracy Metrics
- nDCG@50: 0.436  
- nDCG@100: 0.496  
- Recall@50: 0.800  
- Recall@100: 0.980

##### Latency Distribution
- Mean: 636ms  
- P50 (Median): 613ms  
- P90: 819ms

#### business reports
- **ELO**: 160054.7%  
- **WR**: 301W-230L-19T

##### Accuracy Metrics
- nDCG@50: 0.000  
- nDCG@100: 0.000  
- Recall@50: 0.000  
- Recall@100: 0.000

##### Latency Distribution
- Mean: 599ms  
- P50 (Median): 611ms  
- P90: 727ms

#### FiQa
- **ELO**: 153752.4%  
- **WR**: 288W-253L-9T

##### Accuracy Metrics
- nDCG@50: 0.111  
- nDCG@100: 0.122  
- Recall@50: 0.103  
- Recall@100: 0.135

##### Latency Distribution
- Mean: 639ms  
- P50 (Median): 613ms  
- P90: 819ms

#### MSMARCO
- **ELO**: 147647.8%  
- **WR**: 263W-238L-49T

##### Accuracy Metrics
- nDCG@50: 0.000  
- nDCG@100: 0.000  
- Recall@50: 0.000  
- Recall@100: 0.000

##### Latency Distribution
- Mean: 563ms  
- P50 (Median): 610ms  
- P90: 619ms

#### DBPedia
- **ELO**: 145948.0%  
- **WR**: 264W-243L-43T

##### Accuracy Metrics
- nDCG@50: 0.000  
- nDCG@100: 0.000  
- Recall@50: 0.000  
- Recall@100: 0.000

##### Latency Distribution
- Mean: 587ms  
- P50 (Median): 612ms  
- P90: 656ms

#### PG
- **ELO**: 143454.0%  
- **WR**: 297W-253L-0T

##### Accuracy Metrics
- nDCG@50: 0.000  
- nDCG@100: 0.000  
- Recall@50: 0.000  
- Recall@100: 0.000

##### Latency Distribution
- Mean: 670ms  
- P50 (Median): 615ms  
- P90: 818ms

## Build RAG in Minutes, Not Months
Agentset gives you a complete RAG API with top-ranked rerankers and embedding models built in. Upload your data, call the API, and get accurate results from day one.

### Code Examples
```typescript
import { Agentset } from "agentset";

const agentset = new Agentset();
const ns = agentset.namespace("ns_1234");

const results = await ns.search(
  "What is multi-head attention?"
);

for (const result of results) {
  console.log(result.text);
}
```

## 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.