---
title: naver/v-splade-quality
description: "V-SPLADE is Naver's visual SPLADE model: sparse lexical embeddings over document images and text on a ModernVBERT backbone, bringing interpr. ModernVBERT, 330M parameters."
canonical_url: https://superlinked.com/models/naver-v-splade-quality
last_updated: 2026-08-24
---

# naver/v-splade-quality

V-SPLADE is Naver's visual SPLADE model: sparse lexical embeddings over document images and text on a ModernVBERT backbone, bringing interpretable SPLADE-style retrieval to visual documents

Source: [naver/v-splade-quality on HuggingFace](https://huggingface.co/naver/v-splade-quality)

## Overview

| Field | Value |
|-------|-------|
| Architecture | ModernVBERT |
| Parameters | 330M |
| Tasks | Encode |
| Outputs | Sparse |
| Dimensions | Sparse: 50,368 |
| Max sequence length | 7,999 tokens |
| License | apache-2.0 |
| Inputs | text, image |
| Languages | en |

## Benchmarks

### Vidore3ComputerScienceRetrieval

Domain: technology · Task: retrieval · Language: en

Visual document retrieval on computer science papers and slides

**Quality:** ndcg at 10: 0.5839 · map at 10: 0.4565 · mrr at 10: 0.7488

[Reference](https://arxiv.org/abs/2601.08620)

### Vidore3FinanceEnRetrieval

Domain: finance · Task: retrieval · Language: en

Visual document retrieval on financial reports

**Quality:** ndcg at 10: 0.4651 · map at 10: 0.3504 · mrr at 10: 0.5965

[Reference](https://arxiv.org/abs/2601.08620)

### Vidore3HrRetrieval

Domain: general · Task: retrieval · Language: en

Visual document retrieval on HR-related documents

**Quality:** ndcg at 10: 0.4481 · map at 10: 0.3195 · mrr at 10: 0.5624

[Reference](https://arxiv.org/abs/2601.08620)

### Vidore3PharmaceuticalsRetrieval

Domain: medical · Task: retrieval · Language: en

Visual document retrieval on pharmaceutical documents

**Quality:** ndcg at 10: 0.5324 · map at 10: 0.4196 · mrr at 10: 0.6386

[Reference](https://arxiv.org/abs/2601.08620)