xavier-fuentes/vl-embeddings

VL Embeddings - Qwen3-VL-Embedding-8B

Multimodal embeddings using Qwen/Qwen3-VL-Embedding-8B.

Modes

• —Text + Image: unified multimodal embedding
• —Image only: visual similarity search
• —Text only: text with visual grounding

Matryoshka dimensions

3584, 2048, 1024, 768, 512, 256, 128

When to use this vs Text-Embeddings-8B