Qwen3 Vision-Language Embedding 8B
Qwen3 Vision-Language Embedding 8B is an AI model published by 阿里巴巴, released on 2026-01-08, for embedding模型, with 80.0B parameters, and 32K tokens context length, requiring about 16.32GB storage, under the Apache 2.0 license.
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 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.
Qwen3-VL-Embedding-8B 是 Qwen3-VL-Embedding 系列中的 旗舰级多模态向量模型,面向对检索质量要求极高的场景。模型在多模态理解能力、语义表达精度和跨模态对齐能力上进一步增强,适合用于效果优先的检索与 RAG 系统。
项目说明
模型类型Multimodal Embedding(多模态向量模型)
参数规模8B
网络层数36 层
最大上下文长度32K tokens
向量维度4096(支持 MRL 动态裁剪)
输入模态文本 / 图片 / 截图 / 视频 / 混合模态
指令支持Instruction-aware(支持自定义任务指令)
多语言能力支持 30+ 种语言
量化支持支持低精度量化
许可证Apache 2.0(可商用)
欢迎关注 DataLearner 官方微信,获得最新 AI 技术推送
