Skip to content
TakoAPI
Back to skills

fal-text-to-image

by delorenj

Image & Video Generation
1.5kDownloads reported by clawskills.sh
1Stars
0Views
0 likesSign in to like this skill
—Not yet rated
Your rating

Brief

Generate, remix, and edit images using fal.ai's AI.

Description

What This Skill Does

Generates, remixes, and edits images using fal.ai's AI models from the command line. Supports three modes: text-to-image generation, image-to-image style transfer, and masked inpainting for targeted edits. Automatically selects from models including FLUX, Recraft V3, Imagen4, and Stable Diffusion based on prompt context.

Wraps eight different image generation models under one CLI with automatic model selection, so you get typography-optimized outputs from Recraft and photo-realistic outputs from FLUX without manually choosing endpoints.

When to Use It

  • Generating a product mockup image from a text description
  • Converting a photo to an oil painting or anime style
  • Removing an unwanted object from a photo using inpainting
  • Creating a logo or poster where text must render accurately
  • Replacing a specific region of an image, such as a background or sky

Example Workflow

INPUT

User asks: Generate a minimalist office space, transform it to a cyberpunk style, then add a holographic display on the desk

AGENT
  1. 1

    Run fal-text-to-image with flux-2 using prompt 'Modern office space, minimalist', save as base.png

  2. 2

    Run fal-image-remix on base.png with prompt 'Cyberpunk aesthetic with neon', save as styled.png

  3. 3

    Run fal-image-edit on styled.png with --mask-prompt 'desk' to auto-generate a mask for the desk region

  4. 4

    Apply prompt 'Add holographic display' to the masked area with default strength 0.95

OUTPUT

Three images showing the base office, the cyberpunk-styled version, and the final image with a holographic display added to the desk

Requirements

  • FAL_KEY API key from fal.ai dashboard (https://fal.ai/dashboard/keys)

Installation

clawhub install delorenj/fal-text-to-image

Run the CLI command, or paste the prompt into Claude Code, Cursor, Codex or whichever coding agent you use.

Links