Seed 1.5-VL

Seed1.5-VL, released by ByteDance Seed in May 2025, is a vision-language foundation model designed for advanced general-purpose multimodal understanding and reasoning — combining a dedicated vision encoder with a Mixture-of-Experts language model in an architecture built to handle complex visual reasoning tasks. ByteDance published a comprehensive technical report alongside the release, contributing to the growing body of open research on multimodal model architecture and training methodology.