I Cloned a Stickman Channel Making $9k/mo (Claude + Higgsfield)

Stickly was three weeks old with 13 videos and 15K subscribers, and one video hit 1.6M views in two days. Mac, also three weeks old, had 3M views and 27K subscribers. Even at a low $3 RPM, that’s around $9,000 a month. Both use the simplest style on YouTube: white background, black lines, round-headed stick figures. The whole format can be recreated with Claude Code + Higgsfield in under 20 minutes.

Setup

1. Install Claude Code.
2. Install the Higgsfield CLI: on Higgsfield (50% off) → MCP & CLI → CLI, paste the three lines into Terminal.
3. Connect them (the step most people skip): Claude Code → Customize → Connectors → + → Add custom connector, paste the MCP link, name it Higgsfield, allow.

Pick one lane and a proven topic

Every winning channel picks one topic and sticks to it. My free doc lists the top 30 faceless explainer niches for stickman, each with its RPM.

You don’t need original ideas at first. You need one video to go viral to get monetized, so go after proven topics. Open a reference channel’s Videos → Popular. I used Mac’s When did ancient humans start smoking? (1M views in three weeks).

Rewrite a winning script

Copy the video’s transcript with timestamps (YouTube’s Show transcript, the NexLev extension, or the free TurboScribe). Paste the master script prompt into Claude with your title and the transcript. It studies the structure, hook and pacing, then writes a completely new script.

Visuals with timestamps

Paste the Claude Code prompt from the doc with your script. If it says there are no timestamps, send:

Please add timestamps based on the actual voiceover duration. Use a cinematic YouTube narration pace of 150 words per minute. Each timestamp should represent when that exact line begins. Keep every line separate and unchanged. Do not use fixed intervals.

Approve, and Higgsfield generates the images (watch them live in its image tab). Then:

Download all the images and save them locally using their timestamps as the file names (00:00 → 00-00, 00:10 → 00-10).

Voice that’s yours

Don’t use default ElevenLabs voices that thousands of channels already use. Create voice (design one from a prompt, or clone your own). Split long scripts into sections, switch to v3 and click Enhance for emotional tags like inhales and whispers. Pick the better of the two takes.

Edit in 10 minutes

1. Remove long silences automatically with the free Audacity.
2. In CapCut, drop each image at the time in its file name (00-00 at 0 s, 00-02 at 2 s…). There’s no guessing and no re-listening.
3. Add YouTube Audio Library music, quiet under the voice.

📝 Your notes on this video

Saved on this device only

Copy notesDownload .txt