Cohere Rerank 4 Pro vs Cohere Rerank 3.5

Detailed comparison between Cohere Rerank 4 Pro and Cohere Rerank 3.5. See which reranker best meets your accuracy and performance needs.

Model Comparison

Cohere Rerank 4 Pro takes the lead.

Both Cohere Rerank 4 Pro 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 Cohere Rerank 4 Pro:

  • Cohere Rerank 4 Pro has 175 higher ELO rating
  • Cohere Rerank 4 Pro has a 17.2% higher win rate

Overview

Key metrics

ELO Rating

Overall ranking quality

Cohere Rerank 4 Pro

1627

Cohere Rerank 3.5

1452

Win Rate

Head-to-head performance

Cohere Rerank 4 Pro

58.2%

Cohere Rerank 3.5

41.0%

Accuracy (nDCG@10)

Ranking quality metric

Cohere Rerank 4 Pro

N/A

Cohere Rerank 3.5

0.080

Average Latency

Response time

Cohere Rerank 4 Pro

614ms

Cohere Rerank 3.5

392ms

Visual Performance Analysis

Performance

ELO Rating Comparison

Win/Loss/Tie Breakdown

Latency Distribution (ms)

Breakdown

How the models stack up

MetricCohere Rerank 4 ProCohere Rerank 3.5Description
Overall Performance
ELO Rating
1627
1452
Overall ranking quality based on pairwise comparisons
Win Rate
58.2%
41.0%
Percentage of comparisons won against other models
Pricing & Availability
Price per 1M tokens
$0.050
$0.050
Cost per million tokens processed
Release Date
2025-12-11
2024-12-02
Model release date
Accuracy Metrics
Avg nDCG@10
N/A
0.080
Normalized discounted cumulative gain at position 10
Performance Metrics
Avg Latency
614ms
392ms
Average response time across all datasets

Dataset Performance

By field

Comprehensive comparison of accuracy metrics (nDCG, Recall) and latency percentiles for each benchmark dataset.

MSMARCO

MetricCohere Rerank 4 ProCohere Rerank 3.5Description
Latency Metrics
Mean
458ms
339ms
Average response time
P50
408ms
285ms
50th percentile (median)
P90
615ms
304ms
90th percentile

arguana

MetricCohere Rerank 4 ProCohere Rerank 3.5Description
Latency Metrics
Mean
785ms
570ms
Average response time
P50
768ms
373ms
50th percentile (median)
P90
933ms
617ms
90th percentile

FiQa

MetricCohere Rerank 4 ProCohere Rerank 3.5Description
Latency Metrics
Mean
610ms
364ms
Average response time
P50
585ms
315ms
50th percentile (median)
P90
817ms
401ms
90th percentile

business reports

MetricCohere Rerank 4 ProCohere Rerank 3.5Description
Latency Metrics
Mean
529ms
334ms
Average response time
P50
498ms
293ms
50th percentile (median)
P90
675ms
503ms
90th percentile

PG

MetricCohere Rerank 4 ProCohere Rerank 3.5Description
Latency Metrics
Mean
760ms
458ms
Average response time
P50
720ms
360ms
50th percentile (median)
P90
896ms
615ms
90th percentile

DBPedia

MetricCohere Rerank 4 ProCohere Rerank 3.5Description
Latency Metrics
Mean
541ms
286ms
Average response time
P50
489ms
279ms
50th percentile (median)
P90
729ms
290ms
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.