Skip to main content
Unlisted page
This page is unlisted. Search engines will not index it, and only users having a direct link can access it.

Flux

The Flux integration in ACT3 AI provides access to the Flux generative video model, known for producing smooth, high-fidelity motion sequences from text or image prompts. Flux is particularly effective for stylized cinematic animations, dynamic scene transitions, and image-to-video workflows.

What Flux Does​

Flux generates video clips with smooth, coherent motion that maintains visual consistency across frames. Its physics-aware motion engine handles camera moves, character movement, and environmental effects in a way that feels natural and intentional rather than random.

Key Capabilities​

  • Text-to-Video — Generate short, coherent video clips from descriptive prompts
  • Image-to-Video — Animate a static image by adding motion instructions while preserving its style
  • Physics-Aware Motion — Realistic movement for camera pans, character actions, and environmental effects like wind, water, and fire
  • Style Specialization — Custom artistic filters and motion curves for specific visual looks
  • Extended Clip Mode — Chain multiple generations for longer sequences without losing visual consistency

How to Use​

  1. In the Editor, select a shot or scene and click AI → Flux
  2. Enter your text prompt or upload a still image for animation
  3. Adjust duration, style intensity, and motion settings
  4. Click Generate and review the result in the Preview panel
  5. Accept, refine, or regenerate until the clip matches your vision
  6. Add approved clips to the Timeline or Render Queue

Image-to-Video Workflow​

Flux's Image-to-Video mode is ideal for animating concept art or SDXL-generated still images:

  1. Generate or import a still image
  2. Open it in Flux's Image-to-Video panel
  3. Describe the motion to add: camera movement, character actions, environmental effects
  4. Flux animates the image while preserving its visual style
  5. The result combines your image's aesthetic with smooth, generated motion

This is particularly effective for giving SDXL-generated concept images a motion preview.

Prompt Tips for Flux​

  • Keep prompts under 60 words for Text-to-Video mode
  • Describe motion explicitly and specifically: "slow pan left," "gentle wind moving through trees," "character walks toward camera"
  • For Image-to-Video, describe what should move rather than what should stay still
  • Use Extended Clip Mode for multi-shot sequences that need visual continuity across clips

Credit Usage​

Flux usage is billed via the credit system based on clip length and resolution:

  • Example: 5-second 1080p clip = approximately 3 credits
  • 4K clips cost approximately double
  • Extended Clip Mode charges per segment

Best Use Cases​

  • Cinematic title sequences and motion graphics
  • Adding motion to concept images for animatics
  • Stylized flashbacks or dream sequences
  • Looping animations for social media campaigns
  • Image-based previsualization where an existing asset needs to come to life
  • Projects that need smooth, aesthetically intentional motion over photorealism

When to Use Other Engines​

  • Use Google Veo 3.1 when photorealistic output is the priority
  • Use Runway when stylized or experimental effects are needed
  • Use ComfyUI when chaining Flux with other models in a complex pipeline
  • Use Wan 2.1 for fast collaborative iteration

Best Practices​

  • Save your style and motion settings as a template for visual consistency across a project
  • Combine Image-to-Video with SDXL outputs for unique animated concept previews
  • Use Extended Clip Mode for smooth multi-shot sequences rather than stitching short clips manually
  • Test at low resolution before committing to 1080p or 4K

Troubleshooting​

Motion is too subtle — Increase motion intensity in the Flux settings and describe movement more explicitly in the prompt.

Clip loses visual coherence at the end — Use shorter durations (5–8 seconds) or use Extended Clip Mode for longer sequences.

Image-to-Video changes the original style too much — Reduce style intensity to preserve more of the source image's aesthetic.