---
title: google/siglip2-base-patch16-224
description: SigLIP 2 extends the pretraining objective of SigLIP with prior, independently developed techniques into a unified recipe, for improved sema. SigLIP, 375M parameters.
canonical_url: https://superlinked.com/models/google-siglip2-base-patch16-224
last_updated: 2026-06-05
---

# google/siglip2-base-patch16-224

SigLIP 2 extends the pretraining objective of SigLIP with prior, independently developed techniques into a unified recipe, for improved semantic understanding, localization, and dense features.

Source: [google/siglip2-base-patch16-224 on HuggingFace](https://huggingface.co/google/siglip2-base-patch16-224)

## Overview

| Field | Value |
|-------|-------|
| Architecture | SigLIP |
| Parameters | 375M |
| Tasks | Encode |
| Outputs | Dense |
| Dimensions | Dense: 768 |
| Max sequence length | 64 tokens |
| License | apache-2.0 |
| Inputs | text, image |

## Benchmarks

### 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: 0.8157 · map at 10: 0.7255 · mrr at 10: 0.9302

**Performance (L4 b1 c8):** Corpus 1.6K tok/s · Corpus p50 68.5ms · Query 13.0 mpix/s · Query p50 99.0ms

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