字节跳动刚刚推出了一波新模型: 1️⃣ Doubao 2.1 Pro 面向编程。 2️⃣ Seedance 2.5:4K 视频,每段最长 30 秒。将于 7 月上线,同时推出新的 AI 版权商业化平台。 3️⃣ Seedream 5.0 Pro:图像模型,能处理图像中的文本,支持多层编辑,并输出可编辑的分层设计文件。 4️⃣ Seed-Audio 1.0:零样本多模态音频:一次生成多角色对话、背景音乐和音效。
ByteDance just dropped a wave of new models: 1️⃣ Doubao 2.1 Pro targets coding. 2️⃣ Seedance 2.5: 4K video, up to 30 seconds per clip. Landing in July with a new AI copyright commercialization platform. 3️⃣ Seedream 5.0 Pro: image model that actually handles text in images, multi-layer editing, and outputs editable layered design files. 4️⃣ Seed-Audio 1.0: zero-shot multimodal audio: multi-character dialogue, background music, and sound effects in one pass.