Engineering Visual Intelligence.
Welcome to the official pixPLace technical documentation. This portal is designed to bridge the gap between creative vision and AI mechanics.
Core Workflow
Mode Selection
Choose your medium: Text-to-Image, Photo Editing, Video, or Music.
Model Mapping
Select the algorithm best suited for your task (e.g., Flux for text, Nano for retouching).
Guided Generation
Use reference images and style chips to direct the AI's semantic focus.
Credits & Computation
pixPLace uses a credit-based system reflecting the real-world GPU power required for each task. Your balance is updated in real-time.
💎 SMART BALANCE
Our system automatically calculates the cost before you click "Generate". Top-ups are processed instantly via Telegram Stars.
📊 CONSUMPTION TABLE
Z-Image Turbo
The ultimate tool for Rapid Iteration. Z-Image Turbo uses a specialized single-step distillation process, generating high-quality 1024x1024 visuals in under 3 seconds.
Latency: < 3s
Optimized for brainstorms. Get your first draft almost instantly.
Style Adherence
Highly responsive to Style Chips. Perfect for consistent thematic art.
Nano Banana Pro
Our premium "Image-to-Image" engine. Pro is engineered for Subject Consistency—the ability to take a real person or object and place them in any scenario while keeping them 100% recognizable.
Quad-Reference (1-4 Photos)
Upload up to 4 photos. The AI calculates a 3D semantic map of the subject. Use this for "Digital Photoshoots" where you need the same person in different outfits or locations.
Smart Fidelity
Nano Banana Pro does not use a Strength slider. It uses Automated Subject Isolation—it knows which pixels belong to the face and keeps them sharp while rebuilding the rest of the world.
Flux Pro Advanced
The current State-of-the-Art in AI. Flux Pro excels at Human Anatomy (perfect hands/eyes) and Embedded Text. It is the only model that can reliably "write" on objects within the image.
Text Rendering Guide:
Method:
Put your desired text in double quotes inside the prompt.Example:
"A cyberpunk city with a neon sign that says 'NEO-TOKYO'"Print Maker SDXL
Designed for Graphic Designers and Merch Creators. This is the only model in pixPLace that provides a manual Image Strength slider for granular control over transformation.
Strength 0.1 - 0.3
"Creative Exploration". The AI uses your image only for color palette and basic layout, generating 90% new content.
Strength 0.4 - 0.6
"Style Transfer". Best for taking a pencil sketch and turning it into a professional 3D render or vector illustration.
Strength 0.7 - 0.9
"Quality Restoration". Used to clean up low-res logos, remove artifacts, or add slight textures without changing the design.
Semantic Qwen Edit
The world's first Instruction-Based image editor. It doesn't just re-generate; it understands "verbs" and "objects".
"A man with a blue hat in a park"
Result: AI might change everything to fit the description.
"Change his blue hat to a red crown"
Result: Everything stays same, only the hat changes.
Post-Processing Tools
Alpha Background Removal
Our segmentation AI is trained on studio photography. It identifies the "Subject" vs "Environment" and generates a clean PNG transparency. Ideal for creating stickers or website assets.
Pixel-Perfect Upscaling
Unlike simple bicubic scaling, our HD Upscaler "re-paints" missing details. It uses Generative Super-Resolution to sharpen edges and hallucinate fine textures (like skin pores or fabric weave) that weren't in the original.
Video Generation
Turn static frames into Motion Masterpieces.
Frame Continuity
We use the 1st frame as a strict anchor. The AI calculates "optical flow" to animate hair, water, and camera pans while preserving the original image's look.
Music & Sound (Suno)
Full-spectrum audio generation. From ambient soundscapes to 2-minute vocal tracks. Use our Style Chips to instantly set the BPM, mood, and genre.