@GeekCatX: I've optimized this workflow again. Currently, this should be the lowest-cost and simplest Vox-style video production solution I've run through. The entire workflow has only 4 steps: Step one, use the vox-style skill with Codex to break down scenes and generate storyboard prompts. Step two, …

I’ve optimized this workflow again.

Currently, this should be the cheapest and simplest Vox-style video production solution I’ve ever successfully run.

The entire process only takes 4 steps:

Step 1: Use the vox-style skill with Codex to break down scenes and generate storyboard prompts.

Step 2: Use Codex Image 2 to batch-produce storyboard images.

Step 3: Import the storyboard images into Google Flow, and use Gemini Omni Flash to generate video clips one by one.

Step 4: Download all clips and hand them to ChatCut for one-click editing.

In short:

Split storyboard → Generate images → Convert to video → Auto edit.

No need to design visuals from scratch, nor repeatedly align materials in editing software.

Google AI Pro currently gives 1000 Flow credits per month, which can make producing and testing short videos very low cost.

Of course, how many videos you can actually generate depends on video duration, rework rate, and the model you choose.

But if you want to try AI video at low cost without buying a bunch of expensive tools upfront, this workflow is well worth running.

I’ll post the full process in the Cat Community later, open to members directly.

If you also want to explore AI video, self-media content, and low-cost content production, welcome to join the Cat Community.

We don’t just chase new tools — more importantly, we turn tools into real usable workflows.

知识猫图解 (@GeekCatX): Connected to Google Flow: 15 credits to make a 10s clip, total 45 credits consumed. No edits, direct splicing. Subtitles also handled by skill. There are still some imperfections, feels similar to Veo?