Voyage AI Rerank 2.5 vs Cohere Rerank 3.5
Detailed comparison between Voyage AI Rerank 2.5 and Cohere Rerank 3.5. See which reranker best meets your accuracy and performance needs.
Model Comparison
Voyage AI Rerank 2.5 takes the lead.
Both Voyage AI Rerank 2.5 and Cohere Rerank 3.5 are powerful reranking models designed to improve retrieval quality in RAG applications. However, their performance characteristics differ in important ways.
Why Voyage AI Rerank 2.5:
- Voyage AI Rerank 2.5 has 226 higher ELO rating
- Cohere Rerank 3.5 is 118ms faster on average
- Voyage AI Rerank 2.5 has a 24.8% higher win rate
Overview
Key metrics
ELO Rating
Overall ranking quality
Voyage AI Rerank 2.5
Cohere Rerank 3.5
Win Rate
Head-to-head performance
Voyage AI Rerank 2.5
Cohere Rerank 3.5
Accuracy (nDCG@10)
Ranking quality metric
Voyage AI Rerank 2.5
Cohere Rerank 3.5
Average Latency
Response time
Voyage AI Rerank 2.5
Cohere Rerank 3.5
Visual Performance Analysis
Performance
ELO Rating Comparison
Win/Loss/Tie Breakdown
Accuracy Across Datasets (nDCG@10)
Latency Distribution (ms)
Breakdown
How the models stack up
| Metric | Voyage AI Rerank 2.5 | Cohere Rerank 3.5 | Description |
|---|---|---|---|
| Overall Performance | |||
| ELO Rating | 1629 | 1403 | Overall ranking quality based on pairwise comparisons |
| Win Rate | 62.6% | 37.8% | Percentage of comparisons won against other models |
| Pricing & Availability | |||
| Price per 1M tokens | $0.050 | $0.050 | Cost per million tokens processed |
| Release Date | 2025-08-11 | 2024-12-02 | Model release date |
| Accuracy Metrics | |||
| Avg nDCG@10 | 0.680 | 0.689 | Normalized discounted cumulative gain at position 10 |
| Performance Metrics | |||
| Avg Latency | 610ms | 492ms | Average response time across all datasets |
Dataset Performance
By field
Comprehensive comparison of accuracy metrics (nDCG, Recall) and latency percentiles for each benchmark dataset.
FiQa
| Metric | Voyage AI Rerank 2.5 | Cohere Rerank 3.5 | Description |
|---|---|---|---|
| Accuracy Metrics | |||
| nDCG@5 | 0.108 | 0.121 | Ranking quality at top 5 results |
| nDCG@10 | 0.119 | 0.127 | Ranking quality at top 10 results |
| Recall@5 | 0.098 | 0.118 | % of relevant docs in top 5 |
| Recall@10 | 0.128 | 0.130 | % of relevant docs in top 10 |
| Latency Metrics | |||
| Mean | 540ms | 589ms | Average response time |
| P50 | 564ms | 603ms | 50th percentile (median) |
| P90 | 626ms | 913ms | 90th percentile |
SciFact
| Metric | Voyage AI Rerank 2.5 | Cohere Rerank 3.5 | Description |
|---|---|---|---|
| Accuracy Metrics | |||
| nDCG@5 | 0.865 | 0.868 | Ranking quality at top 5 results |
| nDCG@10 | 0.882 | 0.878 | Ranking quality at top 10 results |
| Recall@5 | 0.892 | 0.887 | % of relevant docs in top 5 |
| Recall@10 | 0.940 | 0.906 | % of relevant docs in top 10 |
| Latency Metrics | |||
| Mean | 667ms | 638ms | Average response time |
| P50 | 621ms | 612ms | 50th percentile (median) |
| P90 | 819ms | 819ms | 90th percentile |
PG
| Metric | Voyage AI Rerank 2.5 | Cohere Rerank 3.5 | Description |
|---|---|---|---|
| Accuracy Metrics | |||
| Latency Metrics | |||
| Mean | 588ms | 661ms | Average response time |
| P50 | 611ms | 613ms | 50th percentile (median) |
| P90 | 746ms | 792ms | 90th percentile |
business reports
| Metric | Voyage AI Rerank 2.5 | Cohere Rerank 3.5 | Description |
|---|---|---|---|
| Accuracy Metrics | |||
| Latency Metrics | |||
| Mean | 714ms | 395ms | Average response time |
| P50 | 617ms | 324ms | 50th percentile (median) |
| P90 | 834ms | 570ms | 90th percentile |
MSMARCO
| Metric | Voyage AI Rerank 2.5 | Cohere Rerank 3.5 | Description |
|---|---|---|---|
| Accuracy Metrics | |||
| nDCG@5 | 0.971 | 0.990 | Ranking quality at top 5 results |
| nDCG@10 | 0.974 | 0.990 | Ranking quality at top 10 results |
| Recall@5 | 0.993 | 1.000 | % of relevant docs in top 5 |
| Recall@10 | 1.000 | 1.000 | % of relevant docs in top 10 |
| Latency Metrics | |||
| Mean | 583ms | 317ms | Average response time |
| P50 | 611ms | 299ms | 50th percentile (median) |
| P90 | 630ms | 320ms | 90th percentile |
DBPedia
| Metric | Voyage AI Rerank 2.5 | Cohere Rerank 3.5 | Description |
|---|---|---|---|
| Accuracy Metrics | |||
| nDCG@5 | 0.657 | 0.691 | Ranking quality at top 5 results |
| nDCG@10 | 0.744 | 0.760 | Ranking quality at top 10 results |
| Recall@5 | 0.062 | 0.063 | % of relevant docs in top 5 |
| Recall@10 | 0.110 | 0.107 | % of relevant docs in top 10 |
| Latency Metrics | |||
| Mean | 571ms | 355ms | Average response time |
| P50 | 610ms | 305ms | 50th percentile (median) |
| P90 | 734ms | 523ms | 90th percentile |
Explore More
Compare more rerankers
See how all reranking models stack up. Compare Cohere, Jina AI, Voyage, ZeRank, and more. View comprehensive benchmarks, compare performance metrics, and find the perfect reranker for your RAG application.