Qwen3-VL-Embedding-8B

Name: Qwen3 Vision-Language Embedding 8B
Author: 阿里巴巴

Embedding modelQwen3

Qwen3 Vision-Language Embedding 8B

Release date: 2026-01-08Updated: 2026-01-08 23:27:57779

Live demoGitHub Hugging Face Compare

Parameters

Context length

32K

Chinese support

Supported

Reasoning ability

Qwen3 Vision-Language Embedding 8B is an AI model published by 阿里巴巴, released on 2026-01-08, for Embedding model, with 8B parameters, and 32K context length, requiring about 16.32GB storage, under the Apache 2.0 license, with a 80.12 score on MMEB-v2-Image.

Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology

Qwen3-VL-Embedding-8B

Model basics

Reasoning traces

Not supported

Thinking modes

Thinking modes not supported

Context length

32K tokens

Max output length

4K tokens

Model type

Embedding model

Modality (in / out)

Text → Embedding

Release date

2026-01-08

Model file size

16.32GB

MoE architecture

Total params / Active params

8B / N/A

Knowledge cutoff

No data

Qwen3-VL-Embedding-8B

Open source & experience

Code license

Apache 2.0

Weights license

Apache 2.0- 免费商用授权

GitHub repo

https://github.com/QwenLM/Qwen3-VL-Embedding/

Hugging Face

https://huggingface.co/Qwen/Qwen3-VL-Embedding-8B

Live demo

No live demo

Qwen3-VL-Embedding-8B

Official resources

Paper

Qwen3-VL-EmbeddingandQwen3-VL-Reranker:AUnifiedFrameworkforState-of-the-ArtMultimodalRetrievalandRanking

DataLearnerAI blog

Qwen3-VL-Embedding-8B

API details

API speed

3/5

No public API pricing yet.

Qwen3-VL-Embedding-8B

Benchmark Results

Qwen3-VL-Embedding-8B currently shows benchmark results led by MMEB-v2-Image (1 / 6, score 80.12). This page also consolidates core specs, context limits, and API pricing so you can evaluate the model from benchmark results and deployment constraints together.