Floniks
One line, endless worlds.
All-modal AI creative canvas. Generate images, produce video and orchestrate end-to-end workflows from a single prompt.
> cat ./overview.md
Floniks is an all-modal AI creative canvas. Type one line; get photoreal stills, cinematic clips, music videos or full short films — without leaving the page.
Beyond one-shot generation, a visual node editor chains 40+ ready-made workflows across image, video and audio, so commerce teams, agencies and indie creators can produce at series scale instead of single assets.
Powered by best-in-class generative models (FAL.ai Nano Banana, MiniMax, Hailuo, APImart) and an MCP integration that lets agents drive the canvas directly from a Claude or Cursor session.
> ls ./features
Text-to-Image Generation
Photoreal to abstract stills from a single prompt, with style controls and reference-image conditioning.
Text- & Image-to-Video
Cinematic camera motion and beat-synced cuts; turn a still or a script into a clip.
Workflow Canvas
XYFlow-based node editor with 40+ preset templates that chain image, video and audio steps into one run.
AI Drama Studio
Script-to-finished-film pipeline with Actor Passport (character consistency) and Scene Passport (location consistency) across shots and episodes.
Music Video Generator
Beat-sync cuts and lyric subtitles auto-aligned to the uploaded track.
Commerce Creative
Bulk marketing-asset generation for e-commerce listings, ads and social — outfit try-on, background swap, restyle, upscale, erase.
VFX Post Studio
Upscale, inpaint, background removal and style transfer for finishing existing footage.
MCP Integration
Drive Floniks from Claude or any MCP-aware agent — prompts in, finished assets out.
> open ./screenshots
> cat ./stack.json
{
"name": "Floniks",
"category": "AI Creative / Multi-Modal",
"stack": [
"Diffusion Models",
"FAL.ai",
"MiniMax / Hailuo",
"XYFlow",
"MCP Integration"
]
}