CivArchive
    [Krea 2 DiT] Nola (znolanola) — Consistent Southern Street Muse - v1.0 - MiniMax H3 LoRA

    # 🔥 NOLA (`znolanola`) — *"Heat Lightning on Elvis Presley Blvd"*

    It is 94 degrees at 11:30 PM in South Memphis, the cicadas are drowning out the box fan in the window, and the neon beer sign across the street is throwing amber stripes across the porch steps.

    **Nola** looks like she stepped straight out of a mid-2000s Southern street film shot on raw 35mm stock—unvarnished, restless, and impossible to look away from. At 5'6" with a lean, lithe frame and narrow shoulders, she wears the heat like a second skin: stretched-out ribbed cotton tanks, faded graphic baby-tees, cheap low-rise nylon shorts, and tarnished silver bamboo hoops that catch the glare of passing headlights. Her skin is fair porcelain with a warm peachy undertone, dusted with natural sun-freckles across the bridge of her nose and damp at the collarbones from the summer humidity.

    Up close, her face tells the whole story before she says a word. She has expressive, almond-shaped hazel-amber eyes with a direct, slightly weary street-smart gaze and a tiny dark beauty mark resting just below the outer corner of her left eye. Her signature look is a lived-in, choppy peroxide-platinum shag with grown-out dark brunette roots and piecey bangs falling across her brow—though she effortlessly switches it up into a messy top-bun pinned with a cheap neon plastic claw clip, tight geometric scalp cornrows, or long waist-length microbraids.

    Whether she is leaning against the chipped doorframe of a cramped hallway in afternoon daylight, slouched in a molded plastic chair under the buzz of a 24-hour fluorescent laundromat, or waiting by a rusted chain-link bus stop in a cropped `"MaisonLatentia"` baby-tee, Nola brings raw, unglamorized documentary soul to every frame.

    ---

    ## ⚡ Quickstart & 1-Click Copy-Paste Prompts

    * **Trigger Token:** `znolanola` (or `znolawoman`)

    * **Recommended LoRA Strength:**

    * **Lustify v10 / Kroma v3 Turbo:** **`1.15 – 1.25`** (`Euler` / `Simple`, `8–10 steps`, `CFG 1.0–1.2`)

    * **Base Krea 2 DiT:** **`0.90 – 1.00`** (`Euler` / `Normal`, `28 steps`, `CFG 3.0`, `Shift 5.0`)

    * **Signature Visual Anchors (Promptable):**

    * *Hair (4 Built-In Styles):* `messy choppy shoulder-length bleach-platinum shag with dark brunette roots` • `messy high top-bun held with a cheap neon plastic claw clip` • `neat geometric cornrows braided along the scalp` • `long waist-length synthetic microbraid extensions`

    * *Face & Skin:* `expressive almond hazel-amber eyes with an honest weary gaze, small dark beauty mark below left outer eye corner, natural dusting of freckles across nose bridge, fair porcelain skin with warm peachy undertone`

    * *Wardrobe:* `fitted faded grey ribbed tank top, baby-blue cotton camisole, cropped "MaisonLatentia" graphic baby-tee, low-rise stretch skirt or nylon shorts, silver hoop earrings`

    ### 📋 Copy-Paste Prompt 1 — Cramped Hallway Daylight Portrait (`896x1152`)

    ```text

    <lora:znolanola_krea2_best:1.15> Raw authentic 85mm documentary daylight close-up portrait of znolawoman (znolanola), showcasing hyper-realistic raw skin pore texture and natural freckles across the nose bridge. Hairstyle: Signature lived-in messy choppy shoulder-length platinum shag with piecey bangs and dark brunette roots. Expressive almond hazel-amber eyes with an honest weary street-smart gaze, small dark beauty mark below left outer eye corner. Wearing a cheap overly tight faded grey ribbed tank top with thin stretched straps, leaning against the chipped doorframe of a cramped inner-city apartment hallway with peeling yellowed paint, dim natural daylight, crisp corneal catchlights, Canon EOS R5 85mm f/2.0.

    ```

    ### 📋 Copy-Paste Prompt 2 — 24-Hour Fluorescent Laundromat (`896x1152`)

    ```text

    <lora:znolanola_krea2_best:1.25> Candid waist-up 50mm documentary portrait of znolawoman (znolanola) slouched in a hard yellow molded-plastic chair inside a fluorescent-lit 24-hour inner-city laundromat late at night. Hairstyle: Hair swept up in an effortless messy high top-bun held with a cheap neon plastic claw clip, with loose wispy tendrils framing her face and damp nape. Wearing a cheap overly tight baby-blue ribbed cotton camisole with raw edges and stretchy black leggings. Rows of scratched industrial washing machines in the background, scuffed linoleum floor, honest tired expression, flat toned stomach, authentic 35mm film grain.

    ```

    ### 📋 Recommended Negative Prompt

    ```text

    blurry, low quality, cgi, 3d render, cartoon, anime, airbrushed, plastic skin, doll, wax figure, deformed iris, extra fingers, bad anatomy, watermark, znolanola text, znola text, misspelled text, gibberish text

    ```

    > 💡 **1-Click Workflow Included:** Every gallery image embeds full A1111/ComfyUI remix metadata—drag any showcase PNG straight into ComfyUI or download `znolanola_krea2_workflow.json` from the files tab.

    ---

    > **Synthetic Character Notice:** `znolanola` is a 100% original fictional persona created from scratch for editorial, narrative, and artistic rendering. All depicted subjects are synthetic adults (21+).

    Description

    MiniMax H3 Core Character LoRA for Nola (znolanola)

    • Release Wave 1 (Core Character LoRA): Designed for both Image-to-Video (I2V) (driving motion from any Nola still frame with locked facial identity) and Text-to-Video (T2V) without needing separate reference-adapter nodes.

    • Trigger Words: znolanola, znolawoman

    • Recommended LoRA Strength: 0.85 – 0.95 (Default: 0.90)

    • Showcase Provenance: All gallery clips on this tab were rendered locally via ComfyUI (MiniMax-H3-FL2VA-Pruned-Q4_K_S.gguf + znolanola_minimax_h3_best.safetensors @ 0.90, no Turbo LoRA), using Nola's high-likeness stills as <Picture 1> I2V start frames (73 frames, 24 fps native audio/video).

    Recommended Sampling Settings

    • Local 8GB ComfyUI (FL2VA Q4_K_S + EasyCache): 15 steps, Euler / Simple, CFG 1.2, SigmaShift 12.0 / 3.0, ImageResizeKJv2 center-crop to 576x768 or 480x832 before MiniMaxH3ImageToVideo (first_frame connected).

    • Cloud Full-Fidelity Native (FL2VA INT8 / BF16): 35 steps, Euler / Simple, CFG 3.5, 124 frames @ 24 fps.

    ComfyUI Installation

    • Place znolanola_minimax_h3_best.safetensors in ComfyUI/models/loras/

    • Drag and drop znolanola_minimax_h3_workflow.json onto your ComfyUI canvas (connect a Nola frame to Load Nola Start Frame for I2V, or bypass first_frame for pure T2V).

    • AutoV2 Hash: f67cc316b5

    FAQ

    LORA
    MiniMax H3

    Details

    Downloads
    64
    Platform
    CivitAI
    Platform Status
    Available
    Created
    11/12/2026
    Updated
    11/13/2026
    Deleted
    -
    Trigger Words:
    znolanola
    znolawoman