Become a founding member.

Andrew AdamsAndrew Adams

What If Factory

Skill1 forks
What If Factory
Run in Claude
Prefer Claude Code? Type /what-if-factory in your agent.

Type one "What if" question. Get a narrated faceless short.

A free alternative to Higgsfield Content Factory. Type one What if question: a story planner writes 3 escalating scenes and a narration, Wan 3.0 makes a clip for each scene, ElevenLabs v4 reads the voiceover, and the Video Editor joins it all with your title as real text, so every letter is exactly what you typed. Every step is its own node, so you can change the idea, the voice or one scene and rerun only that part. Free to copy; running it uses your own Wireflow credits. Our 7 s demo (What if the Moon came closer every night?) cost 104 credits (about 1.04 dollars). Demo: https://cdn.wireflow.ai/videos/remotion/3/4h78m6hgim.mp4

Instructions

# How What If Factory works

Type one "What if..." question and get a short faceless video: a script split into scenes, a clip per scene, a voiceover and a title, joined into one vertical cut. It is a free, open alternative to Higgsfield Content Factory, built from atomic nodes you can check, swap or rerun one at a time.

1. **Write your idea in the text node.** One "What if..." question, the tone (playful, eerie, explainer) and where the story should go.
2. **Story planner (Run any LLM, Claude Sonnet 4.5, 3 cr).** It writes 3 scenes that escalate (one camera shot each, no faces, no words in the picture) and one narration line timed to the clip, as JSON.
3. **JSON Extract Field (free).** One node per scene pulls out that scene's shot, one pulls out the narration.
4. **Video rules + the scene (Prompt Concatenator, free).** The fixed rules say: vertical, one continuous shot, no faces, no text, no letters, no logos.
5. **Wan 3.0, one per scene (about 12 cr a second).** Scene 1 is 2 s, scene 2 is 3 s, scene 3 is 2 s. Sound off: the voiceover carries the audio.
6. **Voiceover (ElevenLabs v4, about 20 cr).** It reads the narration in a stock voice. Pick another voice in the node.
7. **Video Concat (free).** Joins the 3 scenes in order.
8. **Video Editor (free).** Lays the voiceover under the joined clip and sets the title as real text, so every letter is exactly what you typed. **Edit the title here** when you change the idea.

**Demo.** "What if the Moon came a little closer every night?", 7 s, 9:16. No footage or people of ours or anyone else's: everything is generated on this board.

**Cost per run** (catalog prices)

- Story planner: 3 credits (about 0.03 dollars)
- Wan 3.0: about 12 credits a second (about 0.12 dollars); our 2 s, 3 s and 2 s scenes cost about 81 credits together
- Voiceover: about 20 credits (about 0.20 dollars)
- Extract, concat, Video Editor: 0 credits
- Total for a 7 s video: about 104 credits (about 1.04 dollars)

**Tips**

- **Longer video?** Copy a scene row (Extract, Concat, Wan 3.0), ask the planner for one more scene with path scenes.3.shot, wire it into Video Concat, and set the Video Editor frames to the new seconds x 30.
- **A series?** Duplicate the board per episode and change only the idea line.
- **Check the 3 clips before the Video Editor runs.** A bad clip is cheaper to rerun on its own.
- Video models need a paid Wireflow plan; a free account can still copy the board.

See it run

Real outputs from this skill's nodes.

Scene 1 clip, 2 s, 9:16, no sound (Wan 3.0, about 23 cr)
Scene 2 clip, 3 s, 9:16, no sound (Wan 3.0, about 35 cr)
Scene 3 clip, 2 s, 9:16, no sound (Wan 3.0, about 23 cr)
Video Editor: joined scenes + voiceover + your title as real text (7 s = 210 frames; use your seconds x 30, 0 cr). Edit the title here.

More skills

Start here

Ask Claude for your next ad.

Connect once. Ask any time.

  • No credit card
  • You keep the rights to what you make
  • Cancel anytime