A new model in the AI world has been release just two days ago. It is called F-Lite AI which has been released by Freepik and Fal, there is actually two of them F-lite and F-Lite-Texture; they have produced these as open source models but I assume that Lite means there is a Pro model that is not open source.
There is already a Git Repo available for ComfyUI node, that you need to install and then you can use there workflows included. The cool thing is that you just need to run the node and it will download all the models and text encoders etc. If you want to run both models you need about 60GB of storage space.
Once you have installed ComfyUI, you need to install the nodes:
Like all open source models there is a hype and initial curiosity about how good or bad it may be. However I have not be impressed by F-Lite. It’s okay!
The model requires quite large and detailed prompts however with my experiments I have found that even with large details its unable to create and understand the great aspects of the prompt while the quality seems to be substandard of what it produces.
I did all my testing using RunPod with a RTX4090 which was fast using their default workflow provided in the examples folder.
The above custom node of f-lite includes all the needed settings to generate an image, it even has the built in download which will download the models when you first run the workflow. You can switch between the two workflows provided.
Softly lit headshot of a young woman in everyday clothes against a white background, exuding warmth and understated elegance, she is holding a sign that says “F-lite AI”digital artwork showcasing a dashing, elegant man radiating confidence and charm, standing before a luminous neon sign displaying “F-Lite AI” in a sleek, futuristic font. He wears a tailored, modern suit with sharp lines and subtle tech-inspired details, his posture poised and commanding. The neon sign emits a vivid cyan and magenta glow, casting dynamic light across his chiseled features and a high-tech urban backdrop. The scene blends cyberpunk vibrancy with timeless sophistication, with intricate details in his attire, the sign\u2019s glow, and the bustling cityscape, creating a striking, immersive atmosphere.A young woman stands at a bustling city bus stop, her auburn hair catching the late afternoon sun. She’s dressed in a navy coat, scarf loosely draped, clutching a worn novel. The bus stop shelter is glass, streaked with rain from last night\u2019s storm, adorned with a faded timetable and a vibrant ad for a local coffee shop. Behind her, a graffiti-splashed brick wall contrasts with sleek skyscrapers piercing the skyline. Pigeons peck at crumbs near a overflowing trash can, while a street vendor’s cart wafts the scent of roasted chestnuts. A flickering neon sign from a nearby diner reflects on the wet pavement. Commuters rush past, some with umbrellas, others with earbuds, lost in their worlds. A distant siren wails as a double-decker bus approaches, its headlights cutting through the urban haze, promising her ride home.Create a playful digital artwork of a larger-than-life, quirky chef a jovial man with a polka-dot chef’\u2019’s hat and a tie-dye apron grinning as he holds a glowing neon sign reading “F-Lite AI” in a bubbly, cartoonish font. The sign is shaped like a sizzling burger, emitting neon yellow and pink steam that swirls around him. His food truck, parked in a neon-lit cyberpunk alley, has holographic menus and robotic arms flipping pancakes. The backdrop features a starry sky with floating AI drones serving food, blending culinary chaos with futuristic flair in a delightfully quirky scene.A photorealistic 3D render of a spirited nine-year-old girl with a mischievous yet dreamy demeanor, featuring a pair of oversized, bushy fox ears that perk up with curiosity and a delicate, whisker-like pattern of faint silver lines across her cheeks. Her eyes, a luminous amber with flecks of starlight-like gold, shimmer with wonder as she gazes upward. Her hair is a wild cascade of chestnut curls, interwoven with tiny, glowing fireflies that pulse softly, giving her an ethereal glow. She wears a slightly tattered, emerald-green dress with mismatched buttons and a hem embroidered with tiny crescent moons, paired with bare feet dusted with earth. The lighting is a magical blend of moonlight and a faint aurora borealis, casting a soft, prismatic glow that accentuates her freckled, porcelain skin. Her expression is one of quiet awe, lips slightly pursed as if whispering a secret to the stars. The background is a softly blurred night sky over a rolling hill, dotted with glowing dandelions and a distant silhouette of an ancient, gnarled tree. The style balances hyper-realistic textures with a whimsical, storybook charm, emphasizing the interplay of her human features and fox-like traits.
The thing with prompting is that, yes you can write big prompts and essay like descriptions but I think this model requires excessive thesis style prompts which nobody in their right mind would want to write. Even when you do the results are nothing to celebrate (as you can see above).
I think the makers even know that people are not going to write long essays which is why their second example uses the Kijai Node Superprompt to take your brief prompt and try to generate more words (token) for it. Clearly they already know people are not going to like doing this.
Overall I’ve not been impressed, so there are many more new models on the way for us to explore but this one is like the ones you try for an hour and then forget that it exists. At least this version.
If you’d like to support our site please consider buying us a Ko-fi, grab a product or subscribe. Need a faster GPU, get access to fastest GPUs for less than $1 per hour with RunPod.io
