Alibaba NLP Releases Core-Embed 8B for Compositional Multimodal Retrieval

Alibaba NLP has released core-emb-8b, an MLLM-based multimodal embedding model that distills a reranker's compositional judgments to distinguish attribute-object bindings such as "a white plate and a black chair" versus "a black plate and a white chair." The 8B dense embedding model, built on the Qwen3-VL-based VL-Emb backbone, scores 0.666 total average on compositional benchmarks, 5.7 points above its backbone.

It is part of a family that also includes 2B embedding and reranker models.