Now liveThe Skillselion MCP - thousands of ranked skills, loaded into your agent mid-task. No install.Get it →
cloudai-x avatar

World Labs Text Prompt

  • 10 installs
  • 3 repo stars
  • Updated January 22, 2026
  • cloudai-x/world-labs-skills

Generates a 3D world from a natural-language text prompt, with structure for setting, materials, lighting, and environmental context.

About

Covers text-to-world generation in World Labs, including the max length, credits, generation time, and a setting-plus-materials-plus-lighting-plus-context prompt structure. A developer or creator uses it to generate an immersive environment by describing a location.

  • Prompt structure of setting, materials, lighting, and context
  • Max 2,000 characters with plus and mini model timings and costs

World Labs Text Prompt by the numbers

  • 10 all-time installs (skills.sh)
  • Ranked #1,055 of 1,335 Generative Media skills by installs in the Skillselion catalog
  • Data as of Jul 29, 2026 (Skillselion catalog sync)
npx skills add https://github.com/cloudai-x/world-labs-skills --skill world-labs-text-prompt

Add your badge

Show developers this skill is listed on Skillselion. Paste this into your README.

Listed on Skillselion
Installs10
repo stars3
Last updatedJanuary 22, 2026
Repositorycloudai-x/world-labs-skills

What it does

Generates a 3D world from a natural-language text prompt, with structure for setting, materials, lighting, and environmental context.

Files

SKILL.mdMarkdownGitHub ↗

World Labs Text Prompting

Text prompts are the simplest way to generate 3D worlds. Describe a location in natural language, and the model creates an immersive environment.

Quick Reference

  • Max length: 2,000 characters
  • Cost: 1,580 credits (plus) / 230 credits (mini)
  • Generation time: ~5 min (plus) / ~30-45 sec (mini)

Official Example

From World Labs documentation:

A warm, rustic cabin living room with a glowing stone fireplace, cozy leather sofa, wooden beams, and large windows overlooking a snowy forest.

Prompt Structure

Effective prompts describe a location with:

[Setting/Location] + [Materials/Textures] + [Lighting/Atmosphere] + [Environmental Context]

Components

1. Setting: The primary location type (cabin, temple, alley, forest) 2. Materials: Specific textures (rustic wood, polished marble, weathered stone) 3. Lighting: Light sources and quality (glowing, warm, flickering, dusk) 4. Context: Surrounding environment (overlooking forest, narrow corridor)

Style Descriptors

Lighting Keywords

KeywordEffect
glowingSoft emissive light sources
warmCozy, orange-tinted lighting
flickeringDynamic light suggestion
dusk / dawnTransitional daylight
neon-litVibrant artificial lighting
candlelitIntimate, warm glow
moonlitCool, ethereal nighttime

Atmosphere Keywords

KeywordEffect
cozyIntimate, inviting feel
atmosphericMood-heavy, immersive
misty / foggyDepth, mystery
dustyAge, particles
puddlesWet surfaces, reflections
snowyWinter atmosphere

Material Keywords

KeywordEffect
rusticNatural, aged wood
polishedReflective, clean surfaces
weatheredWorn, aged textures
mossyOrganic growth
metallicIndustrial surfaces
leatherFurniture textures

Style Keywords

KeywordEffect
sci-fiFuturistic elements
fantasyMagical, otherworldly
modernContemporary design
vintagePeriod aesthetics
immersiveFull environmental feel

Best Practices

DO

Focus on location: "A cabin living room" not "a picture of a cabin" ✅ Be specific about materials: "glowing stone fireplace" not just "fireplace" ✅ Include spatial context: "large windows overlooking a snowy forest" ✅ Add atmosphere: "warm, cozy, flickering light" ✅ Use concrete details: "wooden beams, leather sofa"

DON'T

Describe people or animals as main subjects ❌ Use vague adjectives: "nice", "beautiful", "amazing" ❌ Request camera angles: handled in Studio, not generation ❌ Contradict styles: "futuristic medieval" confuses the model ❌ Exceed 2,000 characters: prompts are truncated

Example Prompts

Cyberpunk Alley

A cozy sci-fi alley at dusk, flickering neon signs, puddles reflecting colorful lights, narrow corridor between tall buildings, steam rising from vents, cables overhead

Hidden Temple

A mossy ancient temple hidden in a jungle, neon lanterns casting warm pools of light, stone pillars covered in vines, shallow reflecting pool in the center, mysterious and serene atmosphere

Cozy Interior

A warm coffee shop interior on a rainy day, exposed brick walls, wooden tables, rain-streaked windows overlooking a city street, Edison bulbs casting warm light, bookshelves along one wall, leather seating

Natural Landscape

A dramatic coastal cliff at golden hour, crashing waves against ancient rock formations, wild grass swaying in the wind, salt spray catching sunlight, moody clouds on the horizon

API Usage

{
  "model": "Marble 0.1-plus",
  "world_prompt": {
    "type": "text",
    "text_prompt": "A warm, rustic cabin living room with a glowing stone fireplace, cozy leather sofa, wooden beams, and large windows overlooking a snowy forest.",
    "disable_recaption": false
  }
}

Parameters

ParameterTypeDescription
text_promptstringYour location description (max 2,000 chars)
disable_recaptionbooleanIf true, uses your prompt exactly; if false, model may enhance/clarify

Iteration Tips

1. Start with mini: Use Marble 0.1-mini for quick iterations (230 credits, ~30 sec) 2. Refine prompts: Adjust based on results 3. Final generation: Use Marble 0.1-plus for production quality (1,580 credits, ~5 min)

Combining with Images

Text prompts can accompany image inputs to guide interpretation:

{
  "model": "Marble 0.1-plus",
  "world_prompt": {
    "type": "image",
    "image_prompt": {
      "source": "media_asset",
      "media_asset_id": "your-media-asset-id"
    },
    "text_prompt": "Transform into a mystical nighttime scene with aurora borealis"
  }
}

When combined with images:

  • Text guides the model's interpretation of ambiguous areas
  • Can shift lighting, atmosphere, or style
  • If omitted, model auto-generates a caption

Related Skills

  • world-labs-api - API integration details
  • world-labs-image-prompt - Single image input
  • world-labs-studio - Camera animation and recording

Related skills

This week in AI coding

Five minutes, every Monday - the tools, releases and tactics for developers.

unsubscribe anytime.