CivArchive
    Alex Ross FLUX Style - V1
    NSFW
    Preview 39250384
    Preview 39251284
    Preview 39252937
    Preview 39302570
    Preview 39367146
    Preview 39447366
    Preview 39451124
    Preview 39480289
    Preview 40138143
    Preview 40138952

    Alex Ross's art style is celebrated for its hyper-realistic, painterly approach that brings a vivid, almost photographic quality to comic book characters. He uses gouache and watercolor paints to achieve a stunning level of detail, emphasizing realistic lighting, textures, and anatomical precision. Ross’s work often features dramatic, cinematic compositions, with dynamic angles and lifelike expressions that give iconic heroes and villains a sense of grandeur and weight. His characters appear almost tangible, capturing the essence of classic superheroes while grounding them in a realistic, humanized portrayal.

    Ross's distinctive style is also characterized by its use of rich, vibrant colors and a nostalgic aesthetic that evokes the Golden and Silver Age of comics. He often employs a chiaroscuro technique—using stark contrasts of light and shadow to enhance the dramatic atmosphere and emphasize the form of the characters. The result is a timeless, almost mythological presentation of figures like Superman, Batman, and Wonder Woman, blending the epic scale of classic paintings with the dynamic energy of modern comics.

    I wanted to make this lora because of my love for Alex Ross and his work. While its not perfect, it does capture a sense of his style and technique.

    Best Basic Settings:

    1. Sampler: DPM++ 2M OR Euler [Scheduler type of Simple Or Beta]

    2. Sampling Steps: 20

    3. CFG Scale: 3.5

    4. LORA Weight: 0.8-1 [1 seems to work best]

    Description

    FAQ

    LORA
    Flux.1 D

    Details

    Downloads
    223
    Platform
    CivitAI
    Platform Status
    Available
    Created
    11/9/2024
    Updated
    4/30/2026
    Deleted
    -

    Files

    Alex_Ross_FLUX_Style-000001.safetensors

    Mirrors

    1044950_training_data.zip

    Mirrors

    CivitAI (1 mirrors)