deepseek-v4.1POST

DeepSeek V4.1 Flash accepts video and audio in Workbuddy

A user notes V4.1-Flash is multimodal: drop video into Workbuddy for the model to watch, and pair it with Local Whisper to transcribe audio.

Robin | Robotics Data

@Zhibin992Archived 16 hours agoOriginally posted 2026.10.28

Robin | Robotics Data @Zhibin992Originally posted 2026.10.28

Open original on X

告诉大家一个秘密，很多人可能还不知道 DeepSeek V4.1-flash 是视觉模型，也就意味着，你可以直接在 Workbuddy里把视频素材丢给模型，它能看到里面的内容。另外，Workbuddy 本身带一个 Local Whisper 的功能，可以做到录音转写成文本，也就意味着，你可以直接把音频文件丢给它，模型能理解里面的内容。

This post quotes another post: x.com/zhibin992/status/2110524473556324439

x.com/zhibin992/status/2115405337843130766Archived 2026.10.29 12:05