Jina Embeddings v5 Text Small
677M parameter multilingual embedding model built on Qwen3-0.6B-Base with 32K token context length supporting 119+ languages. Features four task-specific LoRA adapters for retrieval, text-matching, clustering, and classification, with Matryoshka learning enabling dimension reduction down to 32. Achieves 67.0 average on MMTEB, best among sub-1B multilingual models. If you want to compare the best embedding models for your data, try Agentset.
Model Information
- Provider
- Jina AI
- License
- CC BY-NC 4.0
- Price per 1M tokens
- $0.050
- Dimensions
- 1024
- Release Date
- 2026-02-18
- Model Name
- jina-embeddings-v5-text-small
- Total Evaluations
- 1260
Performance Record
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
Performance Overview
ELO ratings by dataset
Jina Embeddings v5 Text Small's ELO performance varies across different benchmark datasets, showing its strengths in specific domains.
Jina Embeddings v5 Text Small - ELO by Dataset
Detailed Metrics
Dataset breakdown
Performance metrics across different benchmark datasets, including accuracy and latency percentiles.
business reports
Accuracy Metrics
- nDCG@5
- 0.091
- nDCG@10
- 0.091
- Recall@5
- 0.020
- Recall@10
- 0.024
Latency Distribution
- Mean
- 283ms
- P50 (Median)
- 283ms
- P90
- 312ms
SciFact
Accuracy Metrics
- nDCG@5
- 0.703
- nDCG@10
- 0.734
- Recall@5
- 0.789
- Recall@10
- 0.898
Latency Distribution
- Mean
- 267ms
- P50 (Median)
- 267ms
- P90
- 294ms
PG
Accuracy Metrics
- nDCG@5
- 0.000
- nDCG@10
- 0.000
- Recall@5
- 0.000
- Recall@10
- 0.000
Latency Distribution
- Mean
- 291ms
- P50 (Median)
- 291ms
- P90
- 320ms
FiQa
Accuracy Metrics
- nDCG@5
- 0.838
- nDCG@10
- 0.831
- Recall@5
- 0.677
- Recall@10
- 0.771
Latency Distribution
- Mean
- 300ms
- P50 (Median)
- 300ms
- P90
- 330ms
MSMARCO
Accuracy Metrics
- nDCG@5
- 0.960
- nDCG@10
- 0.954
- Recall@5
- 0.122
- Recall@10
- 0.219
Latency Distribution
- Mean
- 273ms
- P50 (Median)
- 273ms
- P90
- 301ms
arcd
Accuracy Metrics
- nDCG@5
- 0.842
- nDCG@10
- 0.842
- Recall@5
- 0.940
- Recall@10
- 0.940
Latency Distribution
- Mean
- 336ms
- P50 (Median)
- 336ms
- P90
- 370ms
DBPedia
Accuracy Metrics
- nDCG@5
- 0.823
- nDCG@10
- 0.805
- Recall@5
- 0.062
- Recall@10
- 0.123
Latency Distribution
- Mean
- 270ms
- P50 (Median)
- 270ms
- P90
- 297ms
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 Jina Embeddings v5 Text Small with other top embeddings to understand the differences in performance, accuracy, and latency.