- Higgsfield Katana templates use a reference video plus uploaded assets as the creative foundation.
- Preset directions include Personalized Edits, Style Transfer, and Motion Transfer workflows.
- Structured prompts should separate what stays consistent from what changes.
- Iterations cost credits, so refine inputs before rerunning a generation.
- Clean assets (clear character, product, and clothing images) improve output stability.
What a Katana Template Actually Is
A Higgsfield Katana template is a reusable editing structure: a reference video defines movement, framing, and pacing, while your uploaded assets and prompts define the new subject, product, or style. Instead of generating a clip from text alone, Katana rebuilds the reference around your creative direction. The official Katana workflow supports three main preset directions, each suited to a different goal.
| Preset | Category | Best For | Key Input |
|---|---|---|---|
| Personalized Edits | Subject Adaptation | Product ads, personalized social videos | Reference video + replacement subject image |
| Style Transfer | Visual Transformation | Fashion films, brand experiments | Reference video + style direction |
| Motion Transfer | Movement Adaptation | Action videos, performance clips | Reference video + target character/product |
| Product & Social Workflows | Content Scenario | Short-form social campaigns | Reference clip + product or wardrobe assets |
Choose the preset before writing your prompt. A clear editing direction — personalize, restyle, or transfer motion — keeps your instructions focused and reduces wasted generations.
Step-by-Step Template Tutorial
Follow this workflow to run a Katana template from setup to refined result.
Choose a Reference Video
Select footage with clear subject movement and readable camera actions. Match the reference type to your goal: product advertisement, fashion clip, action sequence, or social video. Avoid footage with heavy obstruction or rapid, unreadable cuts.
Upload Your Creative Assets
Add the character, product, or clothing images that should appear in the result. Use clear, well-lit references showing the subject from useful angles, and describe each asset's role so Katana can distinguish it from the reference footage.
Select the Editing Direction
Pick Personalized Edits to adapt the reference to your own subject, Style Transfer to change the visual identity, or Motion Transfer to carry the reference's performance into a new subject.
Describe the Intended Changes
State what must remain consistent (camera rhythm, main action) and what should change (character, product, palette, wardrobe, environment). Keep instructions concrete and free of contradictions.
Generate and Review
Submit the request and check whether the main action, framing, pacing, and replacement subject match your intent. Review credit consumption before each generation.
Refine With Follow-Up Edits
Use conversational modification prompts to adjust one element at a time — clothing, product presentation, camera feel — while explicitly protecting the parts that already work.
A retry or revised generation is a new request and may consume additional credits. Check the displayed consumption before submitting, and refine your inputs instead of repeating identical requests.
Prompt Templates You Can Copy
Effective Katana prompts separate preserved elements from replaced ones. Use these templates as starting structures.
| Template | Purpose | Core Structure |
|---|---|---|
| First Generation | Set full creative direction | Reference role + replacement subject + preserved motion + scene + style |
| Product Replacement | Swap the product | Product details + preserved shot sequence + visibility instructions |
| Character & Wardrobe | New character and clothing | Character description + preserved performance + wardrobe details |
| Style Transfer | Change the look | Style + preserved composition + palette, lighting, mood changes |
| Follow-Up Modification | Targeted refinement | Protect working elements + one focused change |
First Generation Template
"Use the reference video as the foundation. Replace the main subject with [character or product]. Preserve the original movement, camera direction, and pacing. Set the scene in [environment] with [visual style]."
Follow-Up Modification Template
"Keep the current camera movement, pacing, and main action unchanged. Change only [target element]. Make [specific adjustment] while preserving [elements that already work]."
One focused modification at a time produces more stable results than stacking multiple contradictory instructions into a single request.
Input Requirements Checklist
Input quality is the biggest factor in template stability. Prepare assets before uploading.
| Input Type | Requirement |
|---|---|
| Reference video | Clear movement, readable camera actions, supported duration and format |
| Character image | Clean, well-lit, recognizable view; no heavy blur or occlusion |
| Product image | Visible shape, color, logo placement; consistent angles and lighting |
| Clothing image | Full garment visible; silhouette, material, and color readable |
| File links | Accessible without extra login; points to the correct file |
| Rights | Only upload media you own or are authorized to use |
Before You Generate:
- Reference video opens and plays correctly
- Character, product, and clothing images are clear and consistent
- Editing direction (personalize, style, or motion) is chosen
- Prompt separates preserved elements from changes
- Credit balance and expected consumption reviewed
When Katana cannot recognize media, the file is usually inaccessible, unsupported, or incorrectly linked. Test shared links independently and confirm they open without an additional login.
Troubleshooting Template Results
| Problem | Likely Cause | Fix |
|---|---|---|
| Result doesn't match reference | Unclear action, conflicting references | Use one readable movement; simplify instructions |
| Media not recognized | Broken link or unsupported format | Test the file; replace damaged references |
| Subject inconsistent | Vague or conflicting assets | Use one consistent, clear character or product image |
| Credits dropped after retry | Retry counted as new generation | Refine inputs before rerunning; check displayed cost |
| Export unavailable | Generation not finished | Wait for completion, then retry export |
Check identity and motion preservation first, then iterate on instructions. Keep successful parts of the result unchanged when describing the next edit.
FAQ
Q: What is a Higgsfield Katana template?
A reusable editing structure built from a reference video plus your uploaded assets. Katana preserves the reference's movement, framing, and pacing while replacing the subject, product, or style according to your prompt.
Q: Which Katana preset should I choose?
Choose Personalized Edits to adapt a reference to your own subject, Style Transfer to change the visual identity, or Motion Transfer when the reference's movement and performance matter most.
Q: Do template iterations cost extra credits?
Yes. Each retry or revision can be treated as a new generation request and may consume additional credits. Review the displayed consumption before every submission.
Q: Why doesn't my generated video match the reference?
Reference transfer depends on action clarity, subject visibility, image consistency, and instruction specificity. Use a reference with one readable movement, provide clear subject images, and change one thing at a time.