eachlabs-image-generation

What This Skill Does

The eachlabs-image-generation skill is a powerful interface for OpenClaw users to interface directly with the EachLabs Predictions API. It enables the creation of high-quality, professional-grade images from natural language text prompts. By utilizing over 60 diverse AI model families—including the industry-leading Flux series, Google's Imagen and Gemini, and specialized models like Ideogram and Seedream—this skill serves as a comprehensive visual generation hub. It is designed to handle the entire lifecycle of an image request: initiating a prediction, processing parameters like aspect ratio and output format, and monitoring the asynchronous status of the generation task until the final image asset is returned.

Installation

To integrate this skill into your local environment, execute the following command in your terminal: clawhub install openclaw/skills/skills/eftalyurtseven/eachlabs-image-generation

Once installed, ensure you have set your API credentials by setting the EACHLABS_API_KEY environment variable. You can obtain your key by registering at eachlabs.ai.

Use Cases

• Prototyping & Design: Generate instant visual mockups for websites, advertisements, or mobile apps using high-fidelity models like Flux 2 Max.
• Content Creation: Automate the production of social media assets, blog headers, and marketing materials.
• Technical Illustration: Use specialized models like Ideogram V3 Turbo for scenarios where complex text rendering within the image is required.
• Fine-tuned Training: Utilize the dedicated training suite (Z Image Trainer) to create custom LoRA models based on specific subjects or artistic styles.

Example Prompts

1. "Generate a landscape-oriented image using Flux 2 Turbo of a futuristic city skyline at twilight with neon lights."
2. "Create an artistic watercolor-style illustration of a cozy bookstore using the Reve model."
3. "Create a high-quality product shot of a transparent glass bottle on a minimal white background using GPT Image v1.5."

Tips & Limitations

• Asynchronous Nature: Remember that this skill operates via a polling mechanism. If your application or agent script requires immediate confirmation, ensure you implement a loop to check the prediction_id status.
• Model Selection: Different models have specific strengths. For rapid prototyping, use 'Flash' or 'Turbo' variants. For complex, high-detail artistic requests, prioritize 'Max' or 'Pro' models.
• Scope: This skill is strictly for image creation. If your workflow involves modifying, cropping, upscaling, or performing inpainting on an existing file, please use the eachlabs-image-edit skill instead.
• Cost Efficiency: Always select the lightest model possible (like Nano Banana Pro) for simple tasks to minimize latency and potential compute usage.