How to make videos with Codex?
Not in your language? Paste the link or the text into your AI and let it read it for you.
How to do it
Click an avatar to see the source post- 1
For talking-head videos, you can install Yichen's open-source JianYing Skill and hand it to Codex with a single prompt. The author's test: an 8-minute raw clip was automatically cut down to 3 minutes, with segments trimmed down to the millisecond.

- 2
The output is a native JianYing draft, including multiple tracks, subtitles, and volume levels. You can fine-tune it manually in JianYing, or export an MP4 headlessly in the background.

- 3
Doubao voice recognition costs about 0.8 yuan per hour. Some say the Codex token costs can be replaced by using a cheaper model.

- 4
If you don't want to appear on camera, you can use Codex plus Excalidraw to make a hand-drawn whiteboard video from a single sentence. The author says the Skill is open source, and after cloning it, you can produce your first video in ten minutes.

- 5
To replicate a viral video, you can use the open-source Hypit. Give the reference video to Codex, break it into editable components, then swap the host, topic, and language to produce multiple versions in batches.

- 6
There are also two editing Skills to choose from. Pireel focuses on talking-head scenarios and installs with one line in Codex. BaoCut can transcribe, translate, and do simple editing, but it only supports Mac.


What people on X saySorted by usefulness
Everything here comes from real-world tests by users on X. Rules and products change often, so always check the latest. If you know a better way, we'd love to hear it.