Back to all embeddings

Gemini text-embedding-004

Supports 3,000 token context length with task-type specification for retrieval and classification. Legacy model scheduled for deprecation on January 14, 2026, replaced by gemini-embedding-001. If you want to compare the best embedding models for your data, try Agentset.

Leaderboard Rank
#18
of 18
ELO Rating
1366
#18
Win Rate
28.4%
#18
Accuracy (nDCG@10)
0.538
#16
Latency
16ms
#4

Model Information

Provider
Google
License
Proprietary
Price per 1M tokens
$0.020
Dimensions
768
Release Date
2024-05-14
Model Name
text-embedding-004
Total Evaluations
830

Performance Record

Wins236 (28.4%)
Losses541 (65.2%)
Ties53 (6.4%)
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

Gemini text-embedding-004's ELO performance varies across different benchmark datasets, showing its strengths in specific domains.

Gemini text-embedding-004 - ELO by Dataset

Detailed Metrics

Dataset breakdown

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

business reports

ELO 150032.5% WR52W-107L-1T

Accuracy Metrics

nDCG@5
0.000
nDCG@10
0.000
Recall@5
0.000
Recall@10
0.000

Latency Distribution

Mean
15ms
P50 (Median)
15ms
P90
15ms

DBPedia

ELO 150015.0% WR24W-126L-10T

Accuracy Metrics

nDCG@5
0.747
nDCG@10
0.737
Recall@5
0.057
Recall@10
0.108

Latency Distribution

Mean
14ms
P50 (Median)
14ms
P90
14ms

FiQa

ELO 150026.7% WR40W-105L-5T

Accuracy Metrics

nDCG@5
0.744
nDCG@10
0.730
Recall@5
0.647
Recall@10
0.752

Latency Distribution

Mean
16ms
P50 (Median)
16ms
P90
16ms

SciFact

ELO 150043.8% WR70W-83L-7T

Accuracy Metrics

nDCG@5
0.728
nDCG@10
0.729
Recall@5
0.813
Recall@10
0.857

Latency Distribution

Mean
15ms
P50 (Median)
15ms
P90
15ms

MSMARCO

ELO 150030.0% WR48W-90L-22T

Accuracy Metrics

nDCG@5
0.932
nDCG@10
0.918
Recall@5
0.117
Recall@10
0.208

Latency Distribution

Mean
18ms
P50 (Median)
18ms
P90
18ms

ARCD

ELO 15005.0% WR2W-30L-8T

Accuracy Metrics

nDCG@5
0.021
nDCG@10
0.027
Recall@5
0.040
Recall@10
0.060

Latency Distribution

Mean
15ms
P50 (Median)
15ms
P90
15ms

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 Gemini text-embedding-004 with other top embeddings to understand the differences in performance, accuracy, and latency.