CivArchive
    RAD26 - Raw Authenticity Documentary LoRA - v1.0
    NSFW
    Preview 127025992
    Preview 127026148
    Preview 127025986
    Preview 127025993
    Preview 127025976
    Preview 127025991
    Preview 127025989
    Preview 127025977
    Preview 127025981
    Preview 127025983
    Preview 127025979
    Preview 127025975
    Preview 127025988
    Preview 127025984
    Preview 127025982
    Preview 127025987
    Preview 127025980
    Preview 127025990

    https://www.patreon.com/posts/rad26-raw-lora-3-155276180?utm_medium=clipboard_copy&utm_source=copyLink&utm_campaign=postshare_creator&utm_content=join_link

    A LoRA trained with Z Image Turbo at 3000 steps on documentary-style candid photography images.

    Trigger Word: RAD26

    RAD26 captures the defining visual trend of 2026

    Raw Authenticity Documentary (RAD). A direct reaction

    against over-produced AI imagery, RAD26 embraces

    imperfection, mess, genuine emotion and unscripted

    human moments. Every output feels like it was captured

    by a documentary photographer who just happened to be

    in the room.

    This is not fashion. This is not cinematic.

    This is REAL.

    Style characteristics:

    - Candid unposed compositions

    - Natural available light (window light, practical lamps)

    - Authentic film grain texture

    - Messy cluttered real-world environments

    - Genuine unstaged human emotion

    - Raw unretouched skin and imperfect expressions

    - Behind-the-scenes and caught-in-the-moment energy

    - Cinéma vérité documentary framing

    - Wide angle intimate room perspectives

    - Muted desaturated real-world color palette

    Environments covered:

    - Cluttered home interiors

    - Working kitchens and laundry rooms

    - Messy garages and workshops

    - Cramped apartments and hallways

    - Behind-the-scenes creative spaces

    - Chaotic family moments

    - Real working class environments

    - Emotional confrontational scenes

    Human moments covered:

    - Stress and frustration

    - Exhaustion and overwhelm

    - Genuine laughter and connection

    - Grief and emotional breakdown

    - Working and problem solving

    - Caregiving and family tension

    - Quiet solitary moments

    - Multi-generational family scenes

    Best used for:

    - Documentary and photojournalism style content

    - Raw lifestyle brand campaigns

    - Anti-AI aesthetic authentic content

    - Social media real-life storytelling

    - Film and TV production stills

    - Editorial authentic photography

    - 2026 RAD trend content creation

    💡 WHY RAD26 MATTERS IN 2026:

    As AI imagery becomes increasingly polished and

    perfect, the market is swinging hard toward

    authenticity. RAD26 is built for this moment —

    giving creators the tools to generate content

    that feels genuinely human in an AI-saturated world.

    Tips:

    - Always include "RAD26" in your prompt

    - Add "candid, unposed, documentary" for full activation

    - Add "natural window light, film grain" for authentic feel

    - Add "messy environment, cluttered" for real spaces

    - Add "genuine emotion" for authentic expressions

    - AVOID: "beautiful, perfect, cinematic, dramatic lighting"

    - Recommended weight: 0.7–1.0

    Description

    First release. Trained at 750 steps on documentary candid images. Establishes the core RAD26 aesthetic, natural light, film grain, messy environments and genuine human emotion. Lighter style influence, best for subtle documentary grounding or blending with other LoRAs to add authenticity to any scene.
    🔒 WANT ALL 3 VERSIONS? (1500 / 2250 / 3000 steps)

    👉 Get the full RAD26 Pack for $5 on Patreon:

    (your patreon link here)

    This free version (750 steps) is a preview.

    The full pack includes the best quality versions.

    FAQ

    LORA
    ZImageTurbo

    Details

    Downloads
    169
    Platform
    CivitAI
    Platform Status
    Available
    Created
    4/10/2026
    Updated
    4/27/2026
    Deleted
    -

    Files

    RAD26_000000750.safetensors

    Mirrors

    CivitAI (1 mirrors)