Oxi Studios

Oxi Studios is an AI-powered creative platform built for makers, designers, and small studios. It brings together a suite of production-focused tools for generating, converting, and editing images and 3D assets. The main tools include:

• AI Studio: all-in-one image generation and editing hub (Dream, Lumina, Zap, Gem, Mockup, Upscale)
• 3D Relief Generator: converts images into 3D-printable STL bas-relief files
• 3D Model Generator: transforms images into solid 3D GLB models
• AI Vectorizer: converts raster images to clean SVG vector files
• Forge Studio: free SVG to 3MF converter with Figma plugin support
• Silhouette: removes backgrounds and extracts clean silhouettes
• Engrave Cleaner: prepares images for laser engraving
• AI Laser Cut Generator: generates laser-cut SVG files from text prompts
• STL to Depth Map: converts 3D STL files into depth maps for CNC and laser
• Image Prompt Generator: analyses images and generates AI prompts to recreate them

Each tool is powered by a different AI model, chosen for what it does best:

• Seedream 5: high-quality image generation and editing (Dream)
• GPT Image 2: photorealistic image generation with editing support (Lumina)
• Z-Image Turbo: ultra-fast text-to-image generation (Zap)
• Gemini Flash: versatile image creation and editing (Gem)
• Nano Banana: product mockup generation (Mockup)
• Magnific AI: AI-powered image upscaling (Upscale)
• GPT-4 Vision: image analysis for prompt generation
• Depth-Anything V2: depth estimation for 3D relief generation

Each tool processes your input and produces a production-ready output suited to 3D printing, CNC, laser cutting, embroidery, or print workflows.

Oxi Studios uses a credit-based system. Credit costs vary by tool and quality setting:

• Zap (Z-Image Turbo): 1 credit
• Dream (Seedream 5): 1 credit
• Gem (Gemini Flash): 2 credits
• Lumina (GPT Image 2): 1 credit (Low), 2 credits (Medium), 6 credits (High)
• Mockup (Nano Banana): 2 credits (Standard), 5 to 10 credits (Pro)
• Upscale (Magnific AI): 5 to 40 credits depending on the final output size
• AI Vectorizer: 1 credit per conversion
• 3D Relief Generator: 1 credit
• 3D Model Generator: 1 credit
• AI Laser Cut Generator: 5 to 10 credits
• STL to Depth Map: 1 credit
• Image Prompt Generator: 1 credit
• Silhouette and Engrave Cleaner: free, no credits needed

The system always uses your Membership Credits first, then your Casual Credits (purchased credits) after that. Your monthly plan allowance gets used before any credits you've bought separately.

Here's the order:

1. Membership Credits: your monthly plan allowance, used first
2. Casual Credits: purchased top-up credits, used second
3. Combined: if neither pool alone covers the cost but your total does, membership credits are used up first, then the rest comes from casual credits

This way you always get the most out of your monthly plan before dipping into purchased credits.

Oxi Studios offers four plans:

• Free: 10 one-time credits, no credit card needed, access to all tools
• Starter ($10/mo): around 200 credits per month, full commercial use
• Creator ($15/mo): around 300 credits per month, full commercial use
• Pro ($30/mo): around 700 credits per month, full commercial use, great for high-volume work

Not ready to subscribe? You can buy credits as you go. 100 credits for $5.00. All paid plans include full commercial use rights and access to the Figma plugin and Forge Studio.

Yes, you need a free account to access the AI tools. This lets us manage your credits, save your generation history, and provide support. Signing up is quick and only needs basic info. New users get 10 free credits to get started.

Oxi Studios is still being actively built and improved. During this phase we may add new tools, update existing ones, and make changes based on feedback. Beta users get early access to new AI tools and help shape where the platform goes. Your feedback genuinely influences what we build next.

It depends on the tool. Zap is the fastest, usually done in 10 to 30 seconds. Standard generators like Dream, Gem, and Lumina typically take 30 to 60 seconds. More complex jobs like 3D relief generation, upscaling, or laser cut SVG generation can take 1 to 5 minutes. A progress indicator shows while your request is being processed.

Generated images come out as PNG or JPG depending on the tool. For uploads, we accept PNG, JPG, JPEG, and WEBP. The Vectorizer outputs SVG. The 3D Relief Generator outputs STL. The 3D Model Generator outputs GLB. The Laser Cut Generator outputs SVG. All outputs are ready for professional use in manufacturing workflows.

Most uploads are limited to 10MB per file. The 3D Relief Generator accepts files up to 16MB. If your file is too large, try reducing the resolution or compressing it before uploading.

AI Studio is the main image generation and editing hub in Oxi Studios. It combines a canvas-based editor with six AI tools accessible from the left toolbar:

• Dream (Seedream 5): high-quality image generation and editing with Create and Edit modes
• Lumina (GPT Image 2): photorealistic generation with optional reference image editing
• Zap (Z-Image Turbo): ultra-fast text-to-image for quick concepts and drafts
• Gem (Gemini Flash): versatile generation and image editing
• Mockup (Nano Banana): product mockup generation from your images
• Upscale (Magnific AI): AI-powered image upscaling up to 4K

The canvas also has brush, eraser, mask, smudge, pan, and zoom tools, plus AI edit operations like Erase Object, Generative Fill, Remove Background, Replace Background, and Blur Background.

Dream uses Seedream 5 for high-quality image generation. It has two modes:

• Create mode: generate images from a text prompt at 1K, 2K, or 4K resolution in various aspect ratios
• Edit Image mode: upload a reference image and transform it with a prompt. Includes special modes: Clean / Digitize (removes gradients and creates flat-colour designs for embroidery or vector tracing), Line Art (converts to clean lineart for colouring books), and Bas-Relief (creates grayscale depth maps for 3D printing and CNC)

Dream costs 1 credit per generation.

Lumina uses GPT Image 2 and is built for photorealistic, highly detailed image generation. It has three quality levels: Low (1 credit), Medium (2 credits), and High (6 credits), plus three size options (Square 1024x1024, Portrait 1024x1536, Landscape 1536x1024). You can also upload a reference image to edit it with AI. Lumina is better for photorealistic results, while Dream excels at artistic and stylised imagery.

Zap uses Z-Image Turbo for ultra-fast text-to-image generation. It's great for quick concepts, drafts, and rapid iteration when you need results in seconds. Zap costs 1 credit per generation.

Gem uses Gemini Flash and supports both Create mode (text-to-image) and Edit Image mode (upload an image and describe how to change it). It's a versatile general-purpose generator that works well across a wide range of styles. Gem costs 2 credits per generation.

The Mockup tool (powered by Nano Banana) lets you place your product images into realistic scene mockups. Upload up to 14 product images, optionally add a background scene, and describe the placement. It has two tiers: Standard (2 credits, fast and great quality) and Pro (5 to 10 credits, higher resolution at 1K, 2K, or 4K).

The Upscale tool uses Magnific AI to increase image resolution. Upload any image and scale it 2x or 4x. You can choose from multiple Magnific engines (Illusio, Sharpy, Sparkle), optimise for different content types (portraits, art, landscapes, 3D renders, etc.), and adjust creativity, HDR, and resemblance sliders. Credit cost depends on the final output size: 4 credits for 1K to 2K, 8 credits for 2K to 4K.

Once you load an image onto the canvas, you can use these AI edit operations (each costs 1 credit):

• Erase Object: paint a mask over an object to remove it and fill the background
• Generative Fill: paint a mask and describe what to replace it with
• Remove Background: automatically removes the image background
• Replace Background: removes the background and replaces it with a new scene
• Blur Background: blurs the background while keeping the subject sharp

The AI Vectorizer converts raster images (PNG, JPG) into scalable vector graphics (SVG). It's great for logo recreation, sign making, and preparing artwork for large-format printing, vinyl cutting, and CNC. The output SVG works directly with Figma, Illustrator, and other vector tools.

The Vectorizer accepts PNG, JPG, and JPEG files up to 10MB. The output is always an SVG file that can be scaled to any size without losing quality, making it suitable for professional printing and manufacturing.

Our AI Vectorizer uses machine learning to understand image content and create clean vector paths. Traditional auto-trace tools just follow pixel edges, but our AI produces optimised vectors that keep the look and quality of the original image. This works especially well for logos, illustrations, and artwork with complex shapes.

The 3D Relief Generator converts any image into a 3D-printable STL bas-relief file using AI depth estimation. Upload a photo or image and it gets turned into a detailed bas-relief mesh with depth mapping, mesh optimisation, and customisable settings. Great for 3D printing, CNC carving, and laser engraving.

The 3D Relief Generator accepts PNG, JPG, JPEG, GIF, BMP, and WEBP files up to 16MB. The output is an STL file that can be used directly in 3D printing software, CNC programs, or other manufacturing tools.

The tool has four quality levels: Draft, Standard, High, and Ultra. Higher quality produces more detailed meshes but takes longer to process. You can also adjust relief height and base thickness to control how pronounced the 3D effect looks.

Texture Strength: controls how much fine surface detail (like wood grain, scales, or wrinkles) shows up in the relief. Lower values give a smoother result, higher values add more texture.

Depth Map Smoothing: smooths out the overall height map. Higher values give softer, cleaner surfaces. Lower values keep more of the raw AI depth detail.

Texture Smoothing: affects only the surface texture. Higher values soften it, lower values keep it sharper.

Max Height / Min Height: sets the tallest and shortest points of your relief in millimetres, controlling how deep or shallow the carving looks.

Base Thickness: adds a solid base under the relief, which is important for printing stability.

PCA Planar Correction: straightens and corrects perspective so the relief sits flat, useful if the source image has any tilt or distortion.

Yes. The STL files are ready to import into slicing software like OrcaSlicer, BambuStudio, Cura, or PrusaSlicer. The meshes are properly manifold and ready for printing.

The 3D Model Generator turns any image into a solid 3D model and exports it as a GLB file. Unlike the 3D Relief Generator which creates flat bas-relief STL files, this tool generates a full 3D object you can import into Blender, Unity, Unreal Engine, and other 3D software for editing, rendering, or use in games and visualisations.

GLB is the binary version of the GLTF 3D format. It's a widely supported standard that bundles geometry, textures, and materials into a single file. GLB files can be imported into Blender, Unity, Unreal Engine, Sketchfab, and many other 3D applications.

Forge Studio (at forge.oxistudios.com) is a free tool that converts SVG files into print-ready 3MF files. Upload any SVG, preview it in 3D, and export a 3MF file ready for 3D printing. It also has a Figma plugin so you can export any Figma layer directly to 3MF without leaving Figma.

3MF (3D Manufacturing Format) is a modern 3D printing format that supports full colour, materials, and metadata. It's more capable than STL for many workflows and is supported by OrcaSlicer, BambuStudio, PrusaSlicer, Cura, and most modern 3D printing software.

Yes. SVG to 3MF conversion and the Figma plugin are completely free for all users, including those on the Free plan. No credits needed.

The Silhouette tool removes backgrounds from images and extracts clean silhouettes. It's great for laser cutting, vinyl cutting, and preparing artwork where you need a clean subject on a transparent background.

The Engrave Cleaner processes images to prepare them for laser engraving. It removes noise, sharpens edges, and adjusts contrast to produce clean, high-contrast artwork that engraves clearly. This is especially useful for photos or complex images that need simplifying before engraving.

The AI Laser Cut Generator creates laser-cut SVG files from a text description. Describe what you want, choose your material, category, and dimensions, and the tool generates an SVG file ready to cut. Great for creating custom laser cut designs without needing design software. It costs 5 to 10 credits per generation.

The STL to Depth Map tool converts 3D STL files into grayscale depth map images for laser engraving and CNC machining. Upload your STL file and the tool renders a depth map where brightness represents height, ready to use in your laser or CNC software. Processing runs in the background so you can keep browsing while it works. It costs 1 credit per conversion.

The Image Prompt Generator analyses an image and creates a detailed text prompt you can use to recreate something similar in an AI generator. It's useful for reverse-engineering prompts from images you like, or for understanding how to describe a visual style effectively.

Upload an image and our AI (GPT-4 Vision) analyses the visual elements, style, composition, colours, and artistic techniques. It then writes a detailed prompt describing those elements in language that AI image generators understand, so you can recreate similar styles and effects.

Yes. The prompts work with popular AI image generators like Midjourney, DALL-E, Stable Diffusion, and others. You can also use them directly in AI Studio's Dream, Lumina, Zap, or Gem generators.

The tool works with any image, but gets the best results from clear, well-composed images with distinct visual elements. Artwork, photos with a clear subject, and images with a recognisable style tend to produce the most useful prompts.

Uploaded images are stored temporarily on our server only for processing. We don't permanently store, share, or use your images for anything else. Images are deleted after processing is complete. Generated files are also deleted after 24 hours, so download your results when you're done.

Yes. All content is automatically screened to ensure it complies with usage policies. We can't generate images that are harmful, illegal, or inappropriate. This keeps the platform safe and professional for everyone.

Generation can fail for a few reasons: content policy violations, server load, or a technical issue with the AI service. If your request was rejected for content reasons, try rephrasing your prompt. For technical failures, try again or contact support if it keeps happening.

Yes. All paid plans include full commercial use rights. You own what you generate and can use it commercially, including selling products made from 3D-printed reliefs, laser-cut designs, embroidered items, or anything else. The Free plan is for personal and trial use. Check the current terms of service for the full details.

Yes. New users get 10 free credits when they register. This gives you a chance to try out the tools and see the quality before buying more credits or subscribing to a plan.

AI generation doesn't always nail it on the first try. Try adjusting your prompt, changing the quality setting, or switching to a different generator (for example, moving from Zap to Dream or Lumina for better quality). If you're consistently getting poor results, reach out to our support team and we'll help.

You can buy credits through your account at 100 credits for $5.00, no subscription needed. Or subscribe to a monthly plan (Starter, Creator, or Pro) for a better value per credit. Get in touch through the contact page if you need high-volume or custom pricing.

Yes. All uploads and communications are encrypted and we follow standard security practices. Your images are processed securely and deleted after use. We never share your content or use it for training. Your account details are only used to provide our services.

Absolutely. Try adjusting your prompts, experimenting with different generators or quality settings, or using Dream's Clean / Digitize mode to prepare images for specific workflows. If you're still stuck, contact us through the contact page and let us know what you're trying to create.

We're building out tutorials and guides for each tool. Check the How To page for what's available, or reach out if you have specific questions about a workflow or effect.

We're always looking to improve and add new tools based on what users actually need. If you have a suggestion, let us know through the contact page. User feedback genuinely shapes what we build next.

You can reach us through the contact page. Include as much detail as you can about your issue, which tool you were using, what you were trying to do, and any error messages you saw. We'll get back to you as soon as we can.