Google Flow Character Consistency Sheet: 24-Panel Prompt Workflow Guide
Learn how to generate a 24-panel character consistency model sheet in Google Flow from a single reference image using an Identity Lock prompt framework.
AI image and video creator M. Asif (@meAsifAi) has published a production-grade character consistency workflow for Google Flow (@FlowbyGoogle), demonstrating how to anchor a subject's facial identity from a single uploaded reference photo and generate a comprehensive 24-panel model sheet.
Image source: @meAsifAi via X / Google Flow
In generative visual production—whether directing animated sequences, serial graphic novels, AI-assisted short dramas, or game character concepts—one of the most persistent bottlenecks is character drift. As scenes cut across alternating camera perspectives, dynamic emotional expressions, or shifting wardrobe contexts, generative diffusion models frequently warp facial bone structure, alter eye shapes, or modify hairline contours, degrading the visual continuity of the protagonist. To eliminate this friction, M. Asif established a disciplined prompt architecture that treats an initial still photo as an inviolable baseline, synthesizing front portraits, three-quarter angles, full-body silhouettes, localized close-ups, and color palettes onto a unified high-resolution reference board.
The Identity Lock Principle and Facial Feature Preservation
The foundation of this Google Flow workflow is the strict enforcement of an "Identity Lock" directive designed to suppress unwanted model hallucinations, automated beautification filters, and arbitrary character redesigns. Multimodal and diffusion architectures tend to default toward generic, symmetrically idealized facial archetypes when asked to render unseen angles.
The prompt counters this behavioral bias by explicitly declaring the uploaded image as the definitive and exclusive reference for identity. It obligates the model to preserve key anatomical and visual traits with exact fidelity:
- Facial Architecture and Bone Structure: Maintains precise cheekbone prominence, jawline geometry, chin curvature, and facial feature ratios.
- Eyes and Natural Brow Contours: Keeps exact eye geometry, iris pigmentation, eyelid folding, and natural eyebrow thickness without artificial stylization.
- Nose and Lip Proportions: Locks nasal bridge width, tip slope, lip fullness, and mouth contouring.
- Skin Texture and Hair Realism: Preserves authentic skin undertones, natural pore texture, hairline positioning, and original hair length, strand density, and styling.
- Distinctive Physical Markers: Retains natural asymmetry, minor blemishes, or characteristic facial traits that define the individual.
By anchoring the instruction with direct negative constraints—"Do NOT redesign, beautify, reinterpret, or replace the character"—the prompt prevents the model from exerting creative liberties that alter the core likeness of the subject.
Complete 24-Panel Layout Specification and Google Flow Prompt
Borrowing structural conventions from cinematic pre-production turnaround sheets, the prompt arranges 24 distinct diagnostic panels across a single grid. Ranging from macro facial close-ups to four-sided full-body turnaround views, emotional expressions, and fabric references, this composite sheet serves as a permanent anchor for downstream image and video generations.
Below is the full, unabridged prompt shared by creator M. Asif for Google Flow:
Create a professional, ultra-photorealistic CHARACTER CONSISTENCY SHEET using the uploaded reference image as the ONLY source for the character's identity.
IMPORTANT — IDENTITY LOCK:
The uploaded reference image is the definitive identity reference. Reproduce the EXACT SAME PERSON throughout the entire character sheet.
Do NOT redesign, beautify, reinterpret, or replace the character.
Preserve with maximum accuracy:
- exact facial identity
- facial structure and proportions
- eyes, eye shape and iris color
- eyebrows and natural brow shape
- nose shape and proportions
- lips, mouth shape and fullness
- cheeks, jawline and chin
- skin tone and natural skin texture
- age appearance
- hairline
- exact hair color, length, texture and hairstyle
- body proportions and physique
- distinctive facial characteristics
CHARACTER SHEET LAYOUT:
Create one clean, professional reference-board image containing:
1. LARGE FRONT PORTRAIT / FACE CLOSE-UP
2. FRONT VIEW
3. 3/4 LEFT VIEW
4. 3/4 RIGHT VIEW
5. LEFT PROFILE
6. RIGHT PROFILE
7. BACK VIEW
8. FULL-BODY FRONT VIEW
9. FULL-BODY 3/4 VIEW
10. FULL-BODY SIDE VIEW
11. FULL-BODY BACK VIEW
12. NEUTRAL EXPRESSION
13. NATURAL SMILE
14. SOFT SMILE
15. LOOKING LEFT
16. LOOKING RIGHT
17. EYE CLOSE-UP
18. EYEBROW CLOSE-UP
19. NOSE CLOSE-UP
20. LIPS CLOSE-UP
21. HAIR CLOSE-UP
22. ACCESSORIES CLOSE-UP
23. OUTFIT / CLOTHING REFERENCE
24. COLOR PALETTE / SKIN-TONE REFERENCE
CONSISTENCY:
Every panel must show the SAME person.
Keep facial identity, proportions, hairstyle, skin tone and physical characteristics consistent across every angle, pose and expression.
For views not clearly visible in the reference, intelligently reconstruct the character while staying strictly faithful to the visible reference. Do not invent major distinguishing features.
PRESENTATION:
Premium professional film-production character consistency sheet, clean neutral background, studio lighting, razor-sharp details.
The 24 panels are organized into functional diagnostic categories:
- Portraits and Head Turnaround (Panels 1–7): Large front portrait, front view, left/right three-quarter views, full left/right profiles, and rear view establish exact three-dimensional facial depth for rotational camera tracking.
- Full-Body Silhouette and Posture (Panels 8–11): Front, three-quarter, side, and back full-figure perspectives define standing height, shoulder breadth, and overall physical build.
- Expression and Gaze Control (Panels 12–16): Neutral expression, natural smile, soft smile, and left/right directed gazes verify that facial musculature deforms realistically during dialogue or emotive acting.
- Macro Anatomy Close-Ups (Panels 17–21): High-magnification crops of eyes, eyebrows, nose, lips, and hair strands ensure that micro-textures do not dissolve into artifacts during 4K or close-up rendering passes.
- Wardrobe and Palette References (Panels 22–24): Accessories detail, outfit textile references, and skin/fabric color swatch samples allow colorists and lighting engines to balance scene-wide chromatic shifts.
Intelligent Angle Reconstruction and Pipeline Integration
In practical production settings, creators rarely possess multi-angle source footage of their character; most projects begin with a single promotional portrait or user snapshot. Angles such as the back of the head, full lateral silhouettes, and rear wardrobe details are inevitably missing from the input data.
The prompt addresses this constraint in its CONSISTENCY block by directing the model to perform "intelligent reconstruction." Instead of inventing arbitrary traits, tattoos, or extraneous accessories, the AI is instructed to extrapolate missing views using only the visible cues—extending observed hair length, shoulder width, and textile drape with strict logical consistency.
Once generated, this master consistency sheet becomes a cornerstone asset across modern creative workflows:
- AI Video Motion Synthesis: Individual panels or the consolidated sheet can be cropped and fed as multi-image reference inputs into video models like Runway Gen-3, Kling AI, Luma Dream Machine, or Sora, stabilizing temporal character identity during complex camera orbits.
- Visual Novels and Comic Production: Serving as an authoritative reference for ControlNet or IP-Adapter pipelines, the sheet preserves drawing fidelity across hundreds of sequential story panels without manual inpainting overhauls.
- 3D Digital Twin and Game Prototyping: Turnaround angles and neutral lighting provide technical artists with immediate ortho-style references for sculptors, NeRF captures, and Gaussian splatting setups.
For optimal results, input images should feature neutral studio or daylight illumination, sharp focus on facial features, and minimal lens distortion. Supplying a clean baseline ensures that Google Flow can parse the subject's anatomy accurately and project it across all 24 reference perspectives.