发表机构
WeChat Vision, Tencent Inc.（腾讯微信视觉部）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本报告提出WeMM-Embedding多模态嵌入模型家族，含2B、4B、9B变体，经两阶段训练后在公开基准及微信应用中表现优异，已部署并开源。
AI 中文摘要
通用多模态嵌入正成为现代AI系统的核心组件，它能将异构内容表示在共享空间中，用于检索、推荐、分类和智能体系统等应用。本报告提出WeMM-Embedding，这是一个支持文本、图像、视频、视觉文档及任意交错多模态输入的通用多模态嵌入模型家族，具备灵活的输出维度。该家族包含2B、4B和9B三种变体，采用两阶段训练：第一阶段为大规模多模态对齐阶段，第二阶段利用精心整理的数据、细粒度相关性监督及跨尺度知识迁移进行优化。经广泛评估，WeMM-Embedding在多个公开基准上取得领先性能：值得注意的是，2B变体在MMEB-v2上已超越此前领先的8B开源基线，而9B变体则进一步达到80.6的全新最优整体分数。WeMM-Embedding在微信应用中也展现出强劲的实用性能：在包含26个任务的内部基准上取得显著提升，且在14项在线A/B测试中持续改进。目前该模型已大规模部署于推荐和搜索应用，包括微信视频号、公众号、朋友圈及电商服务。我们已发布模型权重和代码，以推动未来研究，链接为this https URL。
英文摘要
Universal multimodal embeddings are becoming a core component of modern AI systems, enabling heterogeneous content to be represented in a shared space for applications such as retrieval, recommendation, classification, and agentic systems. In this report, we present WeMM-Embedding, a family of universal multimodal embedding models supporting text, images, videos, visual documents, and arbitrarily interleaved multimodal inputs with flexible output dimensions. The family comprises 2B, 4B, and 9B variants and is trained in two stages: a large-scale multimodal alignment stage, followed by a refinement stage using curated data, fine-grained relevance supervision, and cross-scale knowledge transfer. Across extensive evaluations, WeMM-Embedding achieves leading performance on multiple public benchmarks. Notably, the 2B variant already surpasses the previously leading 8B open-source baseline on MMEB-v2, while the 9B variant further achieves a new state-of-the-art overall score of 80.6. WeMM-Embedding also demonstrates strong practical performance across WeChat applications, with substantial gains on a 26-task in-house benchmark and consistent improvements across 14 online A/B tests. It has been deployed at scale across recommendation and search applications, including WeChat Channels, Official Accounts, Moments, and e-commerce services. We have released the model weights and code to facilitate future research at https://github.com/Tencent/WeMM-Embedding.
