How to Start a Comic Project Without a Complete Novel (Manual Mode)
You do not need a finished manuscript to start creating AI comic pages. Manual mode lets you write and illustrate one scene at a time, chapter by chapter, as you go. Here is how to use it to build your comic in parallel with your writing.
Who this is for
Most AI comic tools assume you have a finished story. Manual mode is for everyone who does not: writers working chapter by chapter, creators building a story from scratch through visual iteration, and anyone who prefers to see the comic take shape in real time rather than uploading a complete manuscript.
Upload mode vs manual mode: which one to use
Upload a TXT file
You have a complete or near-complete manuscript.
The system automatically splits it into segments and queues them all.
Best for: finished novels, completed web novel arcs, stories you have already written.
Supports resegmentation if you want to adjust the split points.
Manual mode
You write and illustrate scene by scene, as you go.
You control exactly what goes into each segment.
Best for: works in progress, experimental projects, scene testing.
Cannot be resegmented — each segment is added manually, one at a time.
You cannot mix both modes in the same project. Once you start a manual document, the upload option is hidden for that project. If you later complete your manuscript and want to upload the full text, create a new project.
How to start a manual project
Create a new project
Give your project a name that reflects your story or series. Set the art style, quality, and language in the project settings. These apply to every scene you add.
Choose "Start manual creation"
On the project page, instead of uploading a TXT file, click the manual creation option. This creates a virtual document that will hold all your manually added scenes.
Add your first scene
Click "Add segment" and paste or type the text for your first scene (up to 2,000 characters). This does not generate an image yet — it just adds the text to your queue.
Generate the comic for that scene
Click "Create comic" on the segment. The AI simplifies the scene, generates the image, and displays it below the text. Review the result before adding the next scene.
Continue adding scenes
Add the next scene, generate its image, review, and repeat. Each new segment is appended after the last one, building your comic in order.
Writing scenes that generate great images
In manual mode, you control exactly what goes into each segment. This is an advantage — you can write each scene with the comic in mind from the start. Here is what makes a manual segment produce a strong image.
Include a clear setting
Where is the scene? A rooftop at dawn, a crowded market hall, a dimly lit study? The AI needs this to compose the background. Even two or three words of setting description ("moonlit courtyard," "cramped basement office") give the image generator something to work with.
Name the characters and their physical state
Who is in the scene? What are they doing? What do they look like in this moment? "Mira stood at the edge of the bridge, her coat soaked through, gripping the railing with both hands" tells the AI exactly what pose, expression, and costume to render.
Focus each segment on one decisive moment
A segment that tries to cover three scenes will produce a blurry composite image. A segment that zooms in on one moment — a character's face as they read a letter, a sword raised in the moment before the strike — will produce a clear, powerful panel.
Dialogue is fine, but anchor it to action
Dialogue alone can produce good panels, but adding a physical action alongside the words gives the image more to work with. Instead of just the words spoken, add what the character is doing while they speak: "She crossed the room slowly, not looking at him. 'I have been waiting three years for this,' she said."
Using manual mode as a creative process
Some writers find that generating a comic image for each scene as they write it changes how they write. When you see a scene rendered as a panel immediately after writing it, you notice things: a scene that felt vivid in prose produces a flat image because there was no physical action, or a scene you thought was clear turns out to be confusing when compressed to a single visual.
This feedback loop can make the writing better. Scenes that produce weak panels often reveal prose passages that are heavy on interior monologue or abstract description and light on physicality. Adding one concrete visual detail — a gesture, a change in the environment, a physical sensation — usually improves both the prose and the image.
Some creators use this process intentionally: write scene, generate image, revise scene, regenerate image. The comic becomes a visual draft alongside the prose draft, and both improve each other.
Manual mode vs upload: a practical comparison
| Upload TXT | Manual mode | |
|---|---|---|
| Story ready? | Yes, fully written | No, writing as you go |
| Segmentation | Automatic | Manual, full control |
| Best for | Bulk illustration | Scene-by-scene creation |
| Resegment later? | Yes | No |
| Continuous generation? | Yes | Yes (for existing segments) |
Start your comic today, even if the story is not done
Manual mode removes the barrier of needing a finished manuscript. Write the first scene, generate the first panel, and see your story come to life from page one. You build the comic as you build the story — one scene at a time.
Create your first project