V67 Dola‑Doubao

TikTok Aesthetic LoRA – Z Image Turbo - Viral verticals, candid elegance, ByteDance DNA

https://www.dola.com/chat/

Premise: If you use my model and your image is chosen for the CIVITAI cover, you receive 500 Yellow Buzz – even in combination with other models.
R+ images are hidden from the gallery. Trained specifically for Z Image Turbo.

Liking the V67 models is worth more than giving me Buzz (yellow/green).

Overview

Step into TikTok’s most viral aesthetics – cinematic POVs, candid moments, moody lighting, vertical 9:16 composition, soft focus, and natural grain. Inspired by ByteDance’s Dola/Doubao assistant, this LoRA delivers fashionable, modest, well‑tailored characters perfect for short‑form storytelling, transition edits, and “images to videos” workflows.

Trigger Word (place at the beginning)

V67 Dola-Doubao (Dola, Doubao, TikTok:1.1), viral, aesthetic, cinematic, candid, moody, vertical, 9:16, soft focus, natural grain, short videos, images to videos, images to videos

Shorter alternative (quick tests):
V67 Dola-Doubao, TikTok style

Signature Style Elements

• Candid spontaneity – unposed, genuine moments

• Vertical 9:16 framing – native for TikTok

• Moody & cinematic lighting – golden hour, neon night, soft window light

• Soft focus + natural grain – organic smartphone feel

• Modest, fashionable clothing – tailored, elegant, never revealing

• ByteDance DNA – clean, approachable, slightly futuristic Asian aesthetics

• Transition‑ready – poses that work for “images to videos” edits

Trigger Words & Usage

TriggerBest UseV67 Dola-Doubao (Dola, Doubao, TikTok:1.1), viral, aesthetic, cinematic, candid, moody, vertical, 9:16, soft focus, natural grain, short videos, images to videos, images to videosPrimary – full TikTok effectV67 Dola-Doubao, TikTok:1.1, vertical, candidQuick testsDola, Doubao, TikTok aestheticWhen mixing with other LoRAsvertical 9:16, soft focus, natural grainAdd for TikTok finish

Recommendation: Always include vertical, 9:16. For “images to videos” add short videos, images to videos.

Optimized Settings for Z Image Turbo

ParameterValueLoRA Weight0.75 – 0.90CFG Scale2.5 – 3.5Steps4 – 8SamplerDPM++ SDE KarrasResolution720×1280 or 1080×1920 (vertical)

Starting Point Recommendations

GoalLoRA / CFG / StepsSubtle TikTok flavor0.70 / 2.8 / 6Balanced – sweet spot0.80 / 3.0 / 6Strong TikTok dominance (heavy grain, vertical)0.90 / 3.2 / 8“Images to videos” transition ready0.85 / 3.0 / 6

Negative Prompt

text

, low quality, worst quality, ugly, blurry, oversharpened, plastic, artificial, fake, text, watermark, signature, logo, oversaturated, cartoon, 3d render, anime, painting, drawing, flat lighting, harsh flash, unflattering, deformed, distorted, , revealing, , , 

Examples (all with full trigger word)

1. Coffee Shop POV (candid)

Prompt:

text

V67 Dola-Doubao (Dola, Doubao, TikTok:1.1), viral, aesthetic, cinematic, candid, moody, vertical, 9:16, soft focus, natural grain, short videos, images to videos, images to videos – POV of a Dola-like girl looking at her phone in a coffee shop, rain on the window, warm sepia tones, slight smile, messy bun, oversized beige sweater, genuine candid moment

Settings: LoRA 0.80, CFG 3.0, Steps 6, 1080×1920

2. Neon Night Walk (moody)

Prompt:

text

V67 Dola-Doubao (Dola, Doubao, TikTok:1.1), viral, aesthetic, cinematic, candid, moody, vertical, 9:16, soft focus, natural grain, short videos, images to videos, images to videos – Dola-like girl in black leather jacket, walking away from camera, turning head slightly to show profile, neon pink and blue signs reflecting on wet street, modest cyberpunk vibe

Settings: LoRA 0.85, CFG 3.0, Steps 6

3. Before/After Transition (images to videos)

Prompt:

text

V67 Dola-Doubao (Dola, Doubao, TikTok:1.1), viral, aesthetic, cinematic, candid, moody, vertical, 9:16, soft focus, natural grain, short videos, images to videos, images to videos – Split screen: left side messy bun and glasses, right side styled hair and makeup, same Dola avatar, holding a hairbrush, bathroom mirror selfie, ring light reflection, before/after

Settings: LoRA 0.85, CFG 3.2, Steps 8

4. Rooftop Sunset (cinematic)

Prompt:

text

V67 Dola-Doubao (Dola, Doubao, TikTok:1.1), viral, aesthetic, cinematic, candid, moody, vertical, 9:16, soft focus, natural grain, short videos, images to videos, images to videos – Silhouette of a Dola-like woman dancing on a rooftop, arms raised, golden sunset behind, modest flowing dress, city skyline, wind in hair, golden hour

Settings: LoRA 0.80, CFG 3.0, Steps 6

5. Cozy Bedroom Fairy Lights (soft focus)

Prompt:

text

V67 Dola-Doubao (Dola, Doubao, TikTok:1.1), viral, aesthetic, cinematic, candid, moody, vertical, 9:16, soft focus, natural grain, short videos, images to videos, images to videos – Fairy lights behind a Dola-like girl sitting on bed, oversized hoodie, holding phone, looking at camera with soft laugh, messy hair, warm and intimate atmosphere

Settings: LoRA 0.75, CFG 2.8, Steps 6

What This LoRA Excels At (TikTok Context)

• POV first‑person shots (phone in hand, eye contact)

• Candid moments (laughing, looking away, natural expressions)

• Vertical 9:16 compositions (room for text overlays)

• Transition/edit‑friendly poses (images to videos)

• Moody aesthetics (golden hour, blue hour, neon nights)

• Soft focus + grain (analog smartphone feel)

• Modest fashion (tailored, elegant, trend‑aware)

• ByteDance character consistency (recognizable Dola/Doubao traits)

Limitations & Troubleshooting (Z Image Turbo)

IssueSolutionNot enough TikTok/viral feelIncrease LoRA to 0.85+; add viral, candid, POVStatic image, not transition‑readyAdd short videos, images to videos; use action verbs (walking, turning, laughing)Horizontal instead of vertical framingExplicitly add vertical, 9:16 and set resolution to 1080×1920Clothing too revealing or modernStrengthen modest, well-tailored; lower CFG if V67 style is ignoredSoft focus / grain too weakAdd soft focus, natural grain again; try LoRA 0.70 for balanceIncoherent scenesKeep prompt under 100 tokens; reduce LoRA to 0.75; increase CFG to 3.5

Strong Keywords

Style: viral, aesthetic, cinematic, candid, moody, vertical, 9:16, soft focus, natural grain, short videos, images to videos
Character: Dola, Doubao, young woman, Asian features, ByteDance style, approachable, friendly
Clothing: modest, well-tailored, fashionable, coat, sweater, blazer, silk blouse, tailored pants, elegant
Lighting: golden hour, neon, soft window light, warm, teal and orange, rim light
Pose: looking at camera, looking away, laughing, walking, holding phone, candid, over shoulder, POV
Environment: city street, coffee shop, bedroom, rooftop, park, subway, rainy window, fairy lights

Quick Start Workflow

text

Model: Z Image Turbo
LoRA: V67 Dola-Doubao (weight 0.80)
CFG: 3.0 | Steps: 6 | Sampler: DPM++ SDE Karras | Resolution: 1080×1920

Prompt: [FULL TRIGGER WORD] – a Dola-like girl in a beige trench coat, walking through Shibuya crossing at night, lights bokeh, turning back with a slight smile, vertical, candid

Negative: , low quality, plastic, text, watermark, ugly

Credits & Philosophy

“TikTok is not just a platform – it’s a visual language. Candid, vertical, moody, endlessly creative. V67 Dola‑Doubao brings ByteDance’s own AI assistant aesthetic into your generations, ready for the next viral moment.”

Enjoy creating your vertical stories! 🎵📱✨