---
title: Qwen/Qwen3-VL-Embedding-2B
description: The Qwen3-VL-Embedding and Qwen3-VL-Reranker model series are the latest additions to the Qwen family, built upon the recently open-sourced . Qwen3-VL, 2.1B parameters.
canonical_url: https://superlinked.com/models/qwen-qwen3-vl-embedding-2b
last_updated: 2026-09-05
---

# Qwen/Qwen3-VL-Embedding-2B

The Qwen3-VL-Embedding and Qwen3-VL-Reranker model series are the latest additions to the Qwen family, built upon the recently open-sourced and powerful Qwen3-VL foundation model.

Source: [Qwen/Qwen3-VL-Embedding-2B on HuggingFace](https://huggingface.co/Qwen/Qwen3-VL-Embedding-2B)
Base model: [Qwen/Qwen3-VL-2B-Instruct](https://huggingface.co/Qwen/Qwen3-VL-2B-Instruct)

## Overview

| Field | Value |
|-------|-------|
| Architecture | Qwen3-VL |
| Parameters | 2.1B |
| Tasks | Encode |
| Outputs | Dense |
| Dimensions | Dense: 2,048 |
| Max sequence length | 32,768 tokens |
| License | apache-2.0 |
| Inputs | text, image, video |

## Benchmarks

### CQADupstackPhysicsRetrieval

Domain: scientific · Task: retrieval · Language: en

Duplicate question retrieval from StackExchange Physics

Corpus: 38,314 · Queries: 1,039

**Quality:** ndcg at 10: 0.4939 · map at 10: 0.4281 · mrr at 10: 0.4828

[Reference](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/)

### CosQA

Domain: technology · Task: retrieval · Language: en

Code search with natural language queries

Corpus: 6,267 · Queries: 500

**Quality:** ndcg at 10: 0.3448 · map at 10: 0.2677 · mrr at 10: 0.2959

[Reference](https://arxiv.org/abs/2105.13239)

### FiQA2018

Domain: finance · Task: retrieval · Language: en

Financial opinion mining and question answering

Corpus: 57,599 · Queries: 648

**Quality:** ndcg at 10: 0.3793 · map at 10: 0.3051 · mrr at 10: 0.4542

**Performance (L4 b1 c4):** Corpus 494 tok/s · Corpus p50 35.9ms

[Reference](https://sites.google.com/view/fiqa/)

### Flickr30kI2TRetrieval

Domain: general · Task: retrieval · Language: en

Image-to-text retrieval: retrieve captions from images

Corpus: 31,783 · Queries: 1,000

**Quality:** ndcg at 10: 0.8751 · map at 10: 0.8017 · mrr at 10: 0.9653

**Performance (L4 b1 c4):** Corpus 494 tok/s · Corpus p50 35.9ms · Query 0.0 mpix/s · Query p50 4.3s

[Reference](https://shannon.cs.illinois.edu/DenotationGraph/)

### LegalBenchConsumerContractsQA

Domain: legal · Task: retrieval · Language: en

Question answering on consumer contracts

Corpus: 153 · Queries: 396

**Quality:** ndcg at 10: 0.7516 · map at 10: 0.6943 · mrr at 10: 0.6943

[Reference](https://huggingface.co/datasets/nguha/legalbench)

### NFCorpus

Domain: medical · Task: retrieval · Language: en

Biomedical literature search from NutritionFacts.org

Corpus: 3,593 · Queries: 323

**Quality:** ndcg at 10: 0.3689 · map at 10: 0.1432 · mrr at 10: 0.5699

[Reference](https://www.cl.uni-heidelberg.de/statnlpgroup/nfcorpus/)

### SCIDOCS

Domain: scientific · Task: retrieval · Language: en

Citation prediction, document classification, and recommendation for scientific papers

Corpus: 25,656 · Queries: 1,000

**Quality:** ndcg at 10: 0.2090 · map at 10: 0.1250 · mrr at 10: 0.3423

[Reference](https://allenai.org/data/scidocs)

### SciFact

Domain: scientific · Task: retrieval · Language: en

Scientific claim verification using research literature

Corpus: 5,183 · Queries: 300

**Quality:** ndcg at 10: 0.7384 · map at 10: 0.6892 · mrr at 10: 0.6994

[Reference](https://github.com/allenai/scifact)

### StackOverflowQA

Domain: technology · Task: retrieval · Language: en

Programming question answering from Stack Overflow

Corpus: 19,931 · Queries: 1,994

**Quality:** ndcg at 10: 0.9056 · map at 10: 0.8888 · mrr at 10: 0.8888

[Reference](https://arxiv.org/abs/2407.02883)