---
title: 'Claude Code (Free Plan) + YouTube = $77,000/Month'
source: 'https://www.youtube.com/watch?v=WVT2FCjhDDY'
video_id: 'WVT2FCjhDDY'
date: 2026-07-26
duration_sec: 747
---

# Claude Code (Free Plan) + YouTube = $77,000/Month

> Source: [Claude Code (Free Plan) + YouTube = $77,000/Month](https://www.youtube.com/watch?v=WVT2FCjhDDY)

## Summary

This video presents a complete free workflow for creating viral stick figure YouTube videos using AI tools, from script generation to final upload. The creator claims to have built a system using only free plans of Claude AI, 11labs, faziscribe.ai, Google Flow, and a custom Chrome extension, achieving millions of views without any paid subscriptions or design skills.

### Key Points

- **The Stick Man Phenomenon** [00:02] — A stickman channel with simple hand-drawn stick figures gained 7.5 million views with just static images, inspiring the creator to build a free workflow.
- **Single Prompt for Ideas** [01:42] — The creator provides a single optimized prompt for Claude AI (free plan) that generates five viral topic ideas and a complete long-form video script.
- **Critical Workflow Paradigm Shift** [02:35] — The key insight: create voiceover first, then scenes based on natural pauses, not the other way around. This rhythm is what makes videos explode.
- **Voiceover Rhythm Extraction** [03:30] — Using faziscribe.ai, the voiceover is transcribed with exact pause timestamps, which define scene boundaries. This eliminates guessing and syncing issues.
- **Bulk Image Generation with Zappy Flow** [06:47] — The free Zappy Flow Chrome extension automates bulk image generation in Google Flow, processing all scenes at once and downloading them automatically.
- **Final Steps: Editing and Metadata** [10:48] — After matching images to voiceover pauses, Claude generates viral metadata (title, description, tags) and thumbnail prompts for high CTR.

### Conclusion

By following this free workflow, anyone can create viral stick figure videos without paid tools or design skills, leveraging AI for script, voiceover, image generation, and metadata. The key is prioritizing voiceover rhythm over visual synchronization.

## Transcript

sleep. You're scrolling YouTube, thumb moving automatically, not really looking for anything. Um and then one thumbnail stops you cold. A stick man, a simple stops you cold. A stick man, a simple hand-drawn stick man, 7.5 million views.
hand-drawn stick man, 7.5 million views. Um you tap it. No fancy animation, no 3D characters, no cinematic visuals, of stick figures one after another uh stitched together
into a full video. Um that's it. Static images, stick figures with 7.5 million views. You go to the channel, 14 videos, subscribers, 14.5 million total views,
uh with first video published just 2 months ago. You sit up. One thought hits you. If a stick man channel can do this in 2 months, I why can't I? You open your laptop, you try to build
the workflow, but you hit a wall at the very first step. You don't know how to write a script that people actually want to watch. And every AI tool that can help requires a paid subscription. You close the laptop. That stick man
channel just crossed another million views while you're still at zero. I know that feeling exactly cuz I was sitting in the same place uh until I built a complete free workflow that creates these videos from script to
final upload. No paid tools, no design skills, just a system that works. But before we dive in, there's one thing you need. A single prompt, not five tools, [music] not a paid subscription, not a complicated setup, just one prompt.
I spent weeks engineering and optimizing this specifically for this workflow. Uh it's in the pinned comment below. Copy it and now open Claude AI free plan. That's all you need.
You can see I'm on the free plan right now. Paste the prompt and and watch what happens next. The prompt loads. And Claude doesn't just give you one idea, it gives you five. Five viral
topic ideas. Each one engineered to explode if you follow this workflow completely. Pick your number, hit send. What comes back will surprise you. Not just a script, a complete long-form
video script like anywhere from 5 to 15 minutes ready to And it doesn't stop there. Claude automatically creates a text file you can download instantly. Here's the download file. Now, before we move to
I need to stop you. Because what I'm about to tell you next this entire video. Most creators follow this workflow, write script, create text to image prompt for each scenes,
generate scenes, create voice-over, then try to sync everything together. Sounds logical, right? It's completely And if you've been following the same workflow,
this is exactly why your videos aren't exploding. Here's the problem. When you generate scenes first and create the voice-over second, you're forcing two things together that were never built for each other.
creating a collision, and viewers feel that collision even if away. The correct workflow is the opposite. Voice-over first, scene second, always. Here's why.
Look at this voice-over. You can hear the natural pauses between sentences. That pause at 1 second 12 frames. That's where scene one ends. The next pause at 3 seconds 18 frames. Scene two ends there. 6 seconds 7
frames, scene three. 8 seconds 11 frames, scene four. 8 seconds 11 frames, scene four. 12 seconds 16 frames, scene five. Every scene is born from a pause. Not forced, not guessed, not synced
after the fact. Built around the natural rhythm of the voice, this is why that stickman channel crossed 7.5 million views with still images. Not because of fancy tools, not because of better graphics, because of rhythm.
And rhythm starts with the voiceover. Get this right and everything else falls into place. Now let's build the voiceover. Copy your complete script. Now visit 11labs free plan. 10,000 credits every month. This That's
all you need. Go to voices in the left sidebar. Search for Raunak. You'll see it. Viral, relatable, real voice. Add it to your library.
Now go to text-to-speech in the left sidebar. Make sure Raunak is selected. Paste your complete script. Hit generate. Download the audio. Your Now comes the part that separates this workflow from every other tutorial
you've watched. Remember, voiceover first, scene second. But how do you know exactly where each scene should start and end? You're not going to guess. Visit faziscribe.ai.
Link is in the description. Sign up and log in. credit. Once that runs out, create another free account and keep going. Upload your voiceover file. You'll see two modes,
fast and accuracy. Choose accuracy. Every frame matters here. Click transcribe audio and watch what it does. It doesn't just transcribe your words. It finds every pause in your audio.
Every single one with exact timestamps. Pause at two seconds, pause at four seconds, pause at seven seconds. Now you don't need to guess where scenes begin and end. The voice over tells you exactly.
One more thing, you may automatically detect the language of your audio. Hindi, English, Urdu, any language, this workflow works for all of them. Now, download the timestamp script or
simply copy it. Go back to your Claude AI conversation, send. Claude now generates a detailed text to timestamped line. It won't give you everything at once. It
works in parts. When the first batch is done, type next. Keep going until every line has a prompt. Um once all prompts are ready, Claude will ask if you want a text file. Type yes.
Uh within seconds, a complete prompt file is ready to download. Every scene, every timestamp, every prompt Now, it's time to bring these prompts to life. Uh
copy the prompt for scene one, visit flow.google, create a new project. If agent mode is already on, turn it off. Select image mode. Set aspect ratio to
16:9. Set output per prompt to one. Select Nano Banana 2 as your image model. This model gives you the best results uh without hitting generation limits. Paste scene one prompt. Hit send. Within
seconds, scene one is ready. Uh you can repeat this manually for every scene. But, here's the truth. With 100 plus scenes, that process will take hours. There's a faster way, a completely free
faster way. Go to Google and search Zappy Flow Chrome extension. Open the first Chrome Web Store link. You'll see the extension I personally built for you. Completely free and free
forever. One thing, make sure the developer name says zappywala.ai. versions. Add it to Chrome, pin it, launch it.
Now, follow these steps carefully. Go to Google Flow, open a new project, turn agent mode off. You'll see two messages in the extension connected to Google Flow. Agent mode off.
That means everything is ready. Uh now set up the flow for bulk generation, set up the flow for bulk generation, select image mode, set aspect ratio 16:9, set output per prompt one, set model to
nano banana two. Now, paste all your prompts into the prompt box. One line break between each prompt. This is important. You can see we have have prompts with one line break between them. You can simply copy all the
prompts from here or simply upload the prompt text file you downloaded earlier. You can see 103 prompts loaded. One setting before you run, toggle off include serial number in file name. You don't need this for our workflow. Set
your download folder. Click auto save settings and make sure the first option is turned off. Go back to flow, adjust the random delay if you want or leave it as it is. Now, hit generate. Uh and this is where the
magic happens. The extension starts generating every single image and downloading them one by one automatically. And here's the best part, you don't have to watch it. Switch tabs, open another
Chrome profile, minimize the window. It keeps running in the background. No more staring at a screen waiting. Just keep one rule, do not close the tab, the browser, or the extension. Everything else,
you're free. Once complete, you'll see the result. 103 prompts executed, 102 images generated successfully. One failed, scene 15 at the 39-second
timestamp. Don't worry, generate that one manually using the same process. And just like that, all 103 scenes are ready. Now comes the part that ties everything
editing. Open your favorite video editor, import the voiceover file, drag it to the timeline. Now, import all your scene images, drag them to the timeline in the correct
perfectly. Look at the file name of the next scene. Uh it tells you the exact timestamp where it starts. Next scene starts at 4 seconds. Move your playhead to exactly 4 seconds.
Trim from the right. Uh next scene starts at 7 seconds. Move playhead to 7 seconds. Trim again. Sometimes you'll need to adjust the playhead slightly. That's normal. Take your time here. This
is the step that creates the rhythm. This is what made that Stick Man channel cross 7.5 million views. Follow this process for every single scene. Once done, do a full preview. Watch it from the beginning, feel the
rhythm. If it feels right, export the video. description. Now it's time to make it go viral. Um head back to Claude AI. You'll see it's already asking, uh do you want to
generate metadata? Type yes within seconds, uh a viral tags. All generated specifically for your But we still need one more thing, a thumbnail that stops the scroll.
Type this prompt. Ask for five high CTR thumbnail prompts between each. And ask Claude to provide them in a copyable code block. Copy all five prompts, go to the Zappy
extension, toggle it once, paste the prompts, set your download folder, hit generate. Five thumbnails, one click, done. Now go to your YouTube channel, upload your video, use the viral metadata
thumbnail, publish. You just built a complete stick man doodle video from a blank page to a published YouTube video using nothing but free AI tools. No paid subscriptions,
No paid subscriptions, no design skills, no team. that works. If this helped you, hit the like button and subscribe. Small creators like me run on your support. All master prompts
and resources are in the description and pinned comment. And if you want a complete money printing YouTube workflow, you cannot afford to miss the workflow, you cannot afford to miss the video above. See you inside.