Back to all embeddings

Cohere Embed v3

Optimized for English text with 512 token context length supporting retrieval, classification, and clustering tasks. Requires input_type specification distinguishing between search documents and queries for optimal performance. If you want to compare the best embedding models for your data, try Agentset.

Leaderboard Rank
#12
of 14
ELO Rating
1488
#12
Win Rate
41.0%
#10
Accuracy (nDCG@10)
0.686
#13
Latency
7ms
#2

Model Information

Provider
Cohere
License
Proprietary
Price per 1M tokens
$0.100
Dimensions
1024
Release Date
2024-02-07
Model Name
cohere-embed-v3
Total Evaluations
1917

Performance Record

Wins786 (41.0%)
Losses915 (47.7%)
Ties216 (11.3%)
Wins
Losses
Ties

Embedding Models 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 embeddings to manage.

Trusted by teams building production RAG applications

5M+
Documents
1,500+
Teams
99.9%
Uptime

Performance Overview

ELO ratings by dataset

Cohere Embed v3's ELO performance varies across different benchmark datasets, showing its strengths in specific domains.

Cohere Embed v3 - ELO by Dataset

Detailed Metrics

Dataset breakdown

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

MSMARCO

ELO 151947.9% WR115W-82L-43T

Accuracy Metrics

nDCG@5
1.000
nDCG@10
0.996
Recall@5
0.123
Recall@10
0.218

Latency Distribution

Mean
6ms
P50 (Median)
6ms
P90
7ms

PG

ELO 151954.6% WR131W-101L-8T

Latency Distribution

Mean
6ms
P50 (Median)
6ms
P90
7ms

business reports

ELO 151152.1% WR125W-105L-10T

Latency Distribution

Mean
6ms
P50 (Median)
6ms
P90
7ms

SciFact

ELO 150751.2% WR123W-111L-6T

Accuracy Metrics

nDCG@5
0.729
nDCG@10
0.769
Recall@5
0.788
Recall@10
0.900

Latency Distribution

Mean
8ms
P50 (Median)
8ms
P90
9ms

DBPedia

ELO 149644.7% WR106W-110L-21T

Accuracy Metrics

nDCG@5
0.634
nDCG@10
0.619
Recall@5
0.219
Recall@10
0.353

Latency Distribution

Mean
6ms
P50 (Median)
6ms
P90
7ms

NorQuAD

ELO 148924.6% WR59W-75L-106T

Latency Distribution

Mean
9ms
P50 (Median)
9ms
P90
10ms

FiQa

ELO 148344.6% WR107W-130L-3T

Accuracy Metrics

nDCG@5
0.641
nDCG@10
0.650
Recall@5
0.639
Recall@10
0.678

Latency Distribution

Mean
7ms
P50 (Median)
6ms
P90
8ms

ARCD

ELO 13838.3% WR20W-201L-19T

Accuracy Metrics

nDCG@5
0.349
nDCG@10
0.398
Recall@5
0.380
Recall@10
0.520

Latency Distribution

Mean
11ms
P50 (Median)
11ms
P90
12ms

Build RAG in Minutes, Not Months

Agentset gives you a complete RAG API with top-ranked embedding models and smart retrieval built in. Upload your data, call the API, and get accurate results from day one.

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 Cohere Embed v3 with other top embeddings to understand the differences in performance, accuracy, and latency.