Skip to content
Sanskar Sontakke edited this page May 14, 2026 · 1 revision

Onyx Forge Wiki

Welcome to the official wiki for Onyx Forge. This documentation provides in-depth guides on mastering the Forge to create high-performance product advertisements.

image

📖 Table of Contents

  1. Core Concepts
  2. Prompt Engineering
  3. Camera & Optics
  4. Brand Integration
  5. Model Selection Guide

Core Concepts

Onyx Forge is built on the philosophy of "The Forge"—a central workspace where text prompts are synthesized with photographic parameters to create cinematic results. Unlike generic generators, Onyx Forge prioritizes lighting geometry and camera physics.

Prompt Engineering

Auto-Enhance

The "Auto-Enhance" feature uses a specialized LLM layer to inject technical metadata into your prompt. It automatically adds:

  • Volumetric lighting specs
  • Material textures (brushed metal, polished glass, etc.)
  • Environment context

Camera & Optics

Mastering the camera settings is key to the "Onyx Look":

  • Bokeh: Choose 'Strong' or 'Macro' for luxury products to create a shallow depth of field.
  • Lighting Direction: 'Rim' lighting is recommended for sleek tech products to highlight silhouettes.
  • Lens Size: Use '85mm' or '115mm' for distortion-free product portraits.

Brand Integration

To use your own logo:

  1. Upload a high-resolution PNG or SVG in the "Brand Integration" section.
  2. The AI will attempt to place the logo tastefully within the generated environment.
  3. For best results, use logos with transparent backgrounds.

Model Selection Guide

  • Gemini 2.5 Flash Image: Optimized for speed and rapid iteration.
  • Imagen 3.0 Generate: Best for high-fidelity textures and photorealism.
  • Gemini 3.0 Pro Image: The most advanced reasoning, best for complex prompts involving multiple subjects.

For further support, please check the GitHub repository issues.