Popcorn is how Higgsfield handles storyboards and multi-frame sequences. It's Higgsfield's native AI storyboard generator: instead of producing single frames independently, Popcorn creates coherent sequences of up to 8 frames where character identity, lighting, and atmosphere stay consistent across every shot. You can direct each frame yourself in Manual mode or let one prompt expand into a full sequence in Auto mode.
What is Popcorn?
Most image models process each prompt independently: generate a second image with the same character, and the face, lighting, or background shifts. Popcorn generates frames as connected parts of one visual sequence: when you create a character in the first frame, that identity, the lighting logic, and the atmosphere persist across all subsequent frames, and the environment evolves naturally as the scene progresses. Popcorn can also combine real photos and AI-generated images in one sequence.
Open it at higgsfield.ai → Image → Popcorn.
What are the two creation modes?
Manual mode — direct every frame yourself: define the subject, setting, style, and atmosphere shot by shot, with full storyboard-artist control.
Auto mode — write one prompt, choose how many frames you want (up to 8), and Popcorn expands your story into a full cinematic sequence with consistent characters, lighting, and mood.
How do I generate a storyboard?
Choose your input: a written prompt, up to 4 image references (portraits, props, locations), or both combined.
Choose the mode: Manual to control every frame, Auto to generate the full sequence from one prompt.
Generate. Each generation preserves continuity: faces, lighting, and atmosphere stay consistent across all frames. The credit cost is shown on the Generate button before you confirm.
How do references work?
Popcorn accepts up to 4 image references and merges them intelligently: address each by its number in the prompt. For example: "Man from image one in the setting of image three, wearing the outfit from image four."
How do I write a good Popcorn prompt?
Popcorn rewards cinematic thinking. Structure the prompt like a film cue:
Subject — who's on screen
Setting — where the action happens
Style — cinematic, anime, realistic, concept art
Action — what happens
Atmosphere — light, mood, emotion
Example: "Cinematic style. A woman from image one walking through a neon-lit Tokyo street from image three, holding the umbrella from image two. Slow rain, reflections on asphalt, shallow depth of field."
A few things that help: lead with style keywords, describe the action rather than the genre ("a knight walks through fire" works better than "fantasy movie"), use clear emotional language, and refine the text prompt instead of re-uploading inputs when iterating.
Which aspect ratios does Popcorn support?
Ratio | Best for |
|---|---|
3:4 | Portrait or character focus |
2:3 | Cinematic balance |
3:2 | Editorial feel |
1:1 | Social visuals |
9:16 | Reels, TikTok, Shorts |
Keep the input aspect ratio consistent with the target output for the most natural results.
What stays consistent across frames?
Character identity (same face, same body), lighting and clothing, depth, reflections, and spatial logic, and an emotionally coherent atmosphere: every frame reads like a still from the same production.
How do I turn a storyboard into video?
Use Popcorn frames as the visual plan, then animate them with the video models: a frame works as a start frame or reference for Kling or Veo, and chaining the final frame of one clip into the next keeps the motion continuous. See How do I use Kling?