ByteDance’s multimodal foundation models for agents, reasoning, code, media, speech, science, and robotics
Seed is ByteDance’s foundation-model program spanning general-purpose language and vision models, agents, image and video generation, speech, music, scientific AI, and robotics. Seed2.0 offers Pro, Lite, and Mini variants for multimodal reasoning and long-horizon agent workflows, while specialized families such as Seedance and Seedream power media generation. Models are available through ByteDance products and developer APIs, with selected research models released openly.
No reviews yet — be the first to share your experience.
It's easier when you're signed in — Altern helps you get more out of AI.
By continuing you agree to our Terms and Privacy Policy.