Event ·Multi-source

字节跳动豆包大模型2.2推迟发布

字节跳动原计划8月推出的豆包大模型2.2将推迟面世。多位接近字节的人士称,延期是为了通过更充分的预训练和后训练提升编程、工具调用和Agent能力。Seed团队希望模型的Coding能力在年底前达到类似智谱GLM-5.2和Kimi-K3的行业冲击力。字节跳动创始人张一鸣表示公司不会依赖AI蒸馏技术改进模型,CEO梁汝波强调坚持自研,接受短期落后。

Current read

字节推迟豆包2.2发布,专注提升编程和Agent能力,年底目标对标GLM-5.2和Kimi-K3

3 sources58° AI heatLast updated 2026/8/30 13:42:41

Evidence

3 sources

Conflict check

Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.