Cinel AI
Prompts
All promptsBrowse the full libraryPortrait & HeadshotProduct & E-commercePosters, Logos & Graphic DesignInfographics & UI3D Characters & FigurinesFood & DrinkIllustration & AnimeCreative Scenes
AI Image
Free AI Image GeneratorCreate images from text promptsText to Image GeneratorCreate images from text promptsAI Image to Image GeneratorGenerate from reference imagesAI Image EditorEdit existing images with textBackground RemoverRemove backgrounds instantlyImage UpscalerUpscale images 2x or 4xAI UncropExpand images beyond their borders
AI Video
AI Video GeneratorOpen the video generation workspaceText to Video GeneratorCreate videos from text promptsImage to VideoAnimate reference images
AI Models

Image models

KolorsFree Kolors image creation with prompt and reference support.Z-Image TurboFast photoreal generation with bilingual text accuracy.Seedream 4.0Cost-efficient 1K, 2K and 4K visuals with multi-reference editing.Seedream 4.54K generation, text rendering and reference editing.Seedream 5.0 LiteCinematic 3K realism with multi-reference creative direction.Seedream 5 Pro1K fast, 1.5K lightweight, or 2K professional visuals with text and references.GPT Image 2Precise prompts, clean text and polished product shots.GPT Image 2.5Faster GPT Image generation with sharper, consistent edits.Nano Banana 2 LiteFast 1K generation and editing for high-volume iteration.Nano Banana 2Multi-reference edits with fast style control.Nano Banana Pro4K renders and advanced image editing.Nano BananaLightweight Gemini edits for quick changes.

Video models

Seedance 1.0 TurboFast video generation with smooth motion, up to 1080p.Seedance 2.54–30 second video with synchronized audio, 480p or 720p.Seedance 2.0 MiniLower-cost 4–15 second video with synchronized audio, 480p or 720p.Seedance 2.04–15 second video with Fast or Standard mode, up to 1080p.MiniMax H3High-fidelity 4–15 second video with image and reference support.Kling 3.0Strong motion, camera moves and optional sound, up to 4K.Kling 3.0 TurboFaster 3–15 second video at 720p or 1080p, with first-frame input.Kling O1Kling O1 image-to-video with first and optional last frame input.Kling O3Kling O3 text-to-video and first/end-frame generation up to 4K.Grok Imagine Video 1.06–30 second beta video with Fun, Normal and Spicy style modes.Grok Imagine Video 1.51–15 second video with native audio and up to 7 image references.Wan 2.55–10 second text or single-image video with native audio.Wan 2.62–15 second multi-shot video with native audio at 720p or 1080p.Wan 2.72–15 second video with text, first/end frames and up to three image references.Veo 3.1 FastStable Fast Veo 3.1 with 4–8 second video, audio and frame input.Veo 3.1 ProStable Pro Veo 3.1 with 4–8 second video, audio and frame input.Veo 3.1 Fast BetaFast Beta 8-second video with first/end-frame or three-image reference input.Veo 3.1 Pro BetaPro Beta fixed 8-second video with first/end-frame input.Sora 24–12 second video with synchronized audio and one optional first-frame image.
Pricing
-50%Upgrade
  1. Prompts
  2. 3D Characters & Figurines
  3. Tiny Workers Restoring an Antique Pocket Watch

Miniature Craftsmen Restoring an Antique Pocket Watch

{
  "prompt": "Ultra-realistic macro scene depicting  collaboratively repairing an . Several figures delicately clean intricate brass gears using fine brushes, while others stabilize the open watch casing and hold miniature components in place. The pocket watch lies on a textured natural surface, such as , emphasizing age and history. The scene symbolizes time, patience, craftsmanship, and teamwork.",
  "style": {
    "photography_type": "Macro photography",
    "realism": "Ultra-realistic",
    "mood": "Emotional and cinematic",
    "lighting": "Soft, warm, directional lighting with gentle highlights and natural shadows",
    "depth_of_field": "Shallow depth of field with sharp focus on foreground details and smooth background bokeh"
  },
  "composition": {
    "shot_type": "Close-up macro",
    "angle": "Slightly elevated perspective",
    "focus": "Extreme detail on gears, tools, and figurines’ interactions",
    "framing": "Pocket watch centered with workers distributed naturally across the mechanism"
  },
  "textures_and_details": {
    "materials": "Aged metal, polished brass gears, worn leather strap, natural surface grain",
    "details": "Visible dust particles, subtle scratches, fingerprints, and mechanical wear"
  },
  "atmosphere": {
    "tone": "Warm, nostalgic, hopeful",
    "symbolism": "Restoration of time through collective effort and care"
  },
  "render_settings": {
    "aspect_ratio": "4:5",
    "quality": 2,
    "resolution": "High-resolution",
    "style_precision": "Professional studio-level composition"
  }
}

9 credits at 1K. Ratio 4:5.

PhotorealCinematicratio 4x5text only

Details

Category3D Characters & Figurines
Aspect ratio4:5
InputText only
Example generated withGPT Image 2.5
Recommended modelsNano Banana Pro, Nano Banana 2, Seedream 5 Pro
Original author@Just_sharon7 · Original post on X
Original modelnanobanana

Why this prompt works

The tiny-people genre with a gentle story: miniature restorers working together inside an open antique pocket watch, brushing brass gears, steadying the case and carrying parts, as a symbol of patience, craft and the 'restoration of time'. The JSON keeps it photographic with macro lens language, shallow depth of field, warm directional light and wear details like dust, scratches and fingerprints. Our result shows five or six figures in work clothes and caps scattered across the gilded movement on dark weathered wood, one on a ladder, with tweezers and a spare gear nearby. It makes a lovely visual for watch repair shops, heritage brands or anniversary posts.

How to adapt it

VariableWhat it changesExamples
antique pocket watchthe object being repaireda vintage film camera, a gramophone, an old typewriter
tiny restoration workersthe miniature figurestiny engineers in hard hats, tiny clockmakers in aprons
weathered wood or stonesurfacea leather workbench mat, a velvet cloth

Tips

  • Give the tiny workers specific tasks (cleaning, lifting, holding).
  • Wear details (dust, scratches) sell the antique feel.
  • Keep 'shallow depth of field' for a convincing miniature scale.

Related prompts

Glass Skeleton Studying at a Desk 3D Prompt1:1Glass Skeleton Studying at a Desk 3D Prompt@0xROAS
Floating Country-Shaped Island Poster Prompt1:1Floating Country-Shaped Island Poster Prompt@Naiknelofar788
Miniature Construction Site on a Blueprint2:3Miniature Construction Site on a Blueprint@0x00_Krypt
Portrait Holding a Tiny 3D Cartoon Mini-Me2:3Portrait Holding a Tiny 3D Cartoon Mini-Me@Taaruk_
Split Stone and Gold Sculptural Mask PortraitReference1:1Split Stone and Gold Sculptural Mask Portrait@Samann_ai
3D Streetwear Dog Character Design Prompt1:13D Streetwear Dog Character Design Prompt@Taaruk_
Logo Revealed Through Torn Street PostersReference4:5Logo Revealed Through Torn Street Posters@Goodmanprotocol
Isometric Miniature City Diorama on a Vintage Map4:5Isometric Miniature City Diorama on a Vintage Map@Just_sharon7
Silhouette Epic Story Poster (Chinese Prompt)1:1Silhouette Epic Story Poster (Chinese Prompt)@liyue_ai
Lying-Down 'Follow Me' Promo Composite from 2 PhotosReference3:4Lying-Down 'Follow Me' Promo Composite from 2 Photos@BeautyVerse_Lab
Botanical Skincare Pump Bottle Product Shot1:1Botanical Skincare Pump Bottle Product Shot@ChillaiKalan__
Logo as a Frosted Stone Monument on a Mountain PeakReference1:1Logo as a Frosted Stone Monument on a Mountain Peak@zazzygfx
← PreviousGlass Skeleton Studying at a Desk 3D Prompt
Cinel AI

Cinel AI is a free, unlimited AI text-to-image generator with prompt-based editing and AI image-to-video tools. Create images without signing up, edit with GPT Image 2, Seedream 5 Pro, and Nano Banana 2, then turn still frames into video with Seedance 2.5, Veo 3.1, and Kling 3.0.

ProductsFree AI Image GeneratorAI Video GeneratorImage TemplatesWorkspaceText to Image GeneratorImage to ImageText to Video GeneratorImage to VideoAI Image EditorBackground Remover
PromptsAll promptsPortrait & HeadshotProduct & E-commercePosters, Logos & Graphic DesignInfographics & UI3D Characters & FigurinesFood & DrinkIllustration & AnimeCreative ScenesGPT Image 2.5 promptsSources and takedown

AI Image Models

KolorsZ-Image TurboSeedream 4.0Seedream 4.5Seedream 5.0 LiteSeedream 5 ProGPT Image 2GPT Image 2.5Nano Banana 2 LiteNano Banana 2Nano Banana ProNano Banana

AI Video Models

Seedance 1.0 TurboSeedance 2.5Seedance 2.0 MiniSeedance 2.0MiniMax H3Kling 3.0Kling 3.0 TurboKling O1Kling O3Grok Imagine Video 1.0Grok Imagine Video 1.5Wan 2.5Wan 2.6Wan 2.7Veo 3.1 FastVeo 3.1 ProVeo 3.1 Fast BetaVeo 3.1 Pro BetaSora 2
CompanyPricingAboutContactsupport@cinel.aiPrivacyTermsAcceptable Use
© 2026 Cinel AI. All rights reserved.