Voyage 3.5 Lite vs Cohere Embed Multilingual v3

Detailed comparison between Voyage 3.5 Lite and Cohere Embed Multilingual v3. See which embedding best meets your accuracy and performance needs.

Model Comparison

Voyage 3.5 Lite takes the lead.

Both Voyage 3.5 Lite and Cohere Embed Multilingual v3 are powerful embedding models designed to improve retrieval quality in RAG applications. However, their performance characteristics differ in important ways.

Why Voyage 3.5 Lite:

  • Voyage 3.5 Lite delivers better accuracy (nDCG@10: 0.803 vs 0.781)
  • Cohere Embed Multilingual v3 is 12112ms faster on average

Overview

Key metrics

ELO Rating

Overall ranking quality

Voyage 3.5 Lite

1503

Cohere Embed Multilingual v3

1501

Win Rate

Head-to-head performance

Voyage 3.5 Lite

44.4%

Cohere Embed Multilingual v3

42.9%

Accuracy (nDCG@10)

Ranking quality metric

Voyage 3.5 Lite

0.803

Cohere Embed Multilingual v3

0.781

Average Latency

Response time

Voyage 3.5 Lite

36136ms

Cohere Embed Multilingual v3

24024ms

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

MetricVoyage 3.5 LiteCohere Embed Multilingual v3Description
Overall Performance
ELO Rating
1503
1501
Overall ranking quality based on pairwise comparisons
Win Rate
44.4%
42.9%
Percentage of comparisons won against other models
Pricing & Availability
Price per 1M tokens
$0.020
$0.100
Cost per million tokens processed
Release Date
2025-05-20
2024-02-07
Model release date
Accuracy Metrics
Avg nDCG@10
0.803
0.781
Normalized discounted cumulative gain at position 10
Performance Metrics
Avg Latency
36136ms
24024ms
Average response time across all datasets

Dataset Performance

By field

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

PG

MetricVoyage 3.5 LiteCohere Embed Multilingual v3Description
Accuracy Metrics
Latency Metrics
Mean
47120ms
37785ms
Average response time
P50
46178ms
37029ms
50th percentile (median)
P90
54188ms
43453ms
90th percentile

business reports

MetricVoyage 3.5 LiteCohere Embed Multilingual v3Description
Accuracy Metrics
Latency Metrics
Mean
5629ms
5204ms
Average response time
P50
5516ms
5100ms
50th percentile (median)
P90
6473ms
5985ms
90th percentile

DBPedia

MetricVoyage 3.5 LiteCohere Embed Multilingual v3Description
Accuracy Metrics
nDCG@5
0.641
0.619
Ranking quality at top 5 results
nDCG@10
0.632
0.591
Ranking quality at top 10 results
Recall@5
0.219
0.222
% of relevant docs in top 5
Recall@10
0.367
0.329
% of relevant docs in top 10
Latency Metrics
Mean
32706ms
33570ms
Average response time
P50
32052ms
32899ms
50th percentile (median)
P90
37612ms
38606ms
90th percentile

FiQa

MetricVoyage 3.5 LiteCohere Embed Multilingual v3Description
Accuracy Metrics
nDCG@5
0.708
0.637
Ranking quality at top 5 results
nDCG@10
0.736
0.654
Ranking quality at top 10 results
Recall@5
0.687
0.621
% of relevant docs in top 5
Recall@10
0.780
0.692
% of relevant docs in top 10
Latency Metrics
Mean
85143ms
37581ms
Average response time
P50
83440ms
36829ms
50th percentile (median)
P90
97914ms
43218ms
90th percentile

SciFact

MetricVoyage 3.5 LiteCohere Embed Multilingual v3Description
Accuracy Metrics
nDCG@5
0.670
0.723
Ranking quality at top 5 results
nDCG@10
0.719
0.732
Ranking quality at top 10 results
Recall@5
0.718
0.808
% of relevant docs in top 5
Recall@10
0.843
0.833
% of relevant docs in top 10
Latency Metrics
Mean
58245ms
39542ms
Average response time
P50
57080ms
38751ms
50th percentile (median)
P90
66982ms
45473ms
90th percentile

MSMARCO

MetricVoyage 3.5 LiteCohere Embed Multilingual v3Description
Accuracy Metrics
nDCG@5
0.994
0.996
Ranking quality at top 5 results
nDCG@10
0.995
0.994
Ranking quality at top 10 results
Recall@5
0.122
0.122
% of relevant docs in top 5
Recall@10
0.222
0.218
% of relevant docs in top 10
Latency Metrics
Mean
49482ms
32380ms
Average response time
P50
48492ms
31732ms
50th percentile (median)
P90
56904ms
37237ms
90th percentile

NorQuAD

MetricVoyage 3.5 LiteCohere Embed Multilingual v3Description
Accuracy Metrics
Latency Metrics
Mean
7081ms
3824ms
Average response time
P50
6939ms
3748ms
50th percentile (median)
P90
8143ms
4398ms
90th percentile

ARCD

MetricVoyage 3.5 LiteCohere Embed Multilingual v3Description
Accuracy Metrics
nDCG@5
0.928
0.925
Ranking quality at top 5 results
nDCG@10
0.935
0.933
Ranking quality at top 10 results
Recall@5
0.960
0.940
% of relevant docs in top 5
Recall@10
0.980
0.960
% of relevant docs in top 10
Latency Metrics
Mean
3679ms
2305ms
Average response time
P50
3605ms
2259ms
50th percentile (median)
P90
4231ms
2651ms
90th percentile

Explore More

Compare more embeddings

See how all embedding models stack up. Compare OpenAI, Cohere, Jina AI, Voyage, and more. View comprehensive benchmarks, compare performance metrics, and find the perfect embedding for your RAG application.