Upload or generate your 3D model with AI, fine-tune your scene in real time, change materials and go live with an immersive high-fidelity 3D-AR viewer in seconds.
Drag & drop
3D Model
All your pages, models and live links, together in one place.
Explore the architecture, workflow, and core capabilities of the BASE 3D-AR Studio
BASE opens straight into the 3D Generator. It is the first item in the left menu and the default view, so the first thing you see is the AI generator: the viewer with your model strip along the bottom and the generation panel on the right (a sheet on phones), where you drop an image or describe a model to build it. From there the menu runs 3D Generator · Model Editor · Page Designer · Library · Analytics · Resources, and switching between them is instant with no page reload. Your account, plan and settings open from the account pill at the bottom of the sidebar. Already have a GLB? Use the upload button in the top bar, then fine-tune it in the Model Editor.
The 3D engine prioritizes WebGPU for faster rendering and advanced shader support, with automatic fallback to WebGL 2.0 on unsupported browsers. Tone mapping uses ACESFilmic for cinema-grade color accuracy.
The engine renders in sync with your display's refresh rate (VSync) with no artificial frame cap. On 60 / 90 / 120 Hz screens it runs at the matching frame rate, supporting smooth playback up to 120 fps on capable hardware.
If sustained frame drops are detected, the engine makes a one-time conservative adjustment (render scale and shadow resolution) to keep motion smooth. Full quality can be restored anytime.
Models are cached locally via basebros-3d-cache-v1. Repeat visits load instantly with zero bandwidth cost. Assets are served from a global CDN for minimal latency worldwide.
.GLB (Binary glTF) is the recommended format for optimal performance. Max file size: 50 MB. Drag & drop multiple files for batch upload.
Three steps from 3D file to live AR experience:
Full MeshPhysicalMaterial pipeline with support for Base Color, Roughness, Metalness, Normal, Emissive, and AO maps. Edit every surface channel in real-time with texture upload.
Fine-tune viewer behavior with animation playback, auto-rotate speed, zoom toggle with zoom in/out limits, shadow controls, model positioning offsets, and camera orbit limits.
Variance Shadow Mapping at 2048×2048 resolution with adjustable softness. Shadow blur samples auto-calculate based on radius for optimal quality/performance balance.
Independent settings for desktop and mobile. Logo position, AR button style, and layout are configured per device with live dual-mockup preview.
Built-in AI-powered assistant trained on Studio documentation. Get instant answers, workflow guidance, and troubleshooting help, available 24/7 in the sidebar. Pop it out into a floating window to keep it open while you work. It reads your live account context (plan, capacity, Live models, This month, page count), so answers are specific to you. If a reply ever fails, your typed message is kept and you get a retry instead of an empty bubble.
Click the + button next to the assistant's message box to attach up to 4 images per message, each under 5MB. Any image type your browser reports is accepted, plus HEIC/HEIF from iPhone (recognised by file extension and converted automatically). The + button hides once you reach 4. Thumbnails appear above the input, each with an X to remove it.
The assistant actually looks at what you send, so the fastest way to report a visual problem is to screenshot the screen you are stuck on and ask about it. Attached images are used for that reply only and are not kept in your chat history. The same + button works in Support messages and Bug Reports.
Your scene, material and control edits stay as a local preview until you click Save. A status card shows "Unsaved changes" with Save and Discard buttons, and tells you what is pending: Model, Page or Model + Page. The Model Editor and Page Designer share this one card, and a single Save commits the model first, then the page design, straight to your live public page. There is no separate publishing step to remember. Discard asks for confirmation and returns everything to the last saved state.
Asset pipeline, model cards, and model organization tools.
Drag & drop one or multiple .GLB files to upload, up to 5 files per batch. Max file size is 50 MB per model. Every upload is optimized automatically: geometry is DRACO-compressed right in your browser during upload, and textures are resized to a maximum of 2048 px and served as WebP. The system tracks file size, upload date, and generates metadata automatically. Live models depend on your plan; when they are full the upload card shows your options. Every upload counts as a new model this month.
/quick-view opens a GLB instantly without signing in and without uploading: drop a .GLB (max 50 MB) and it is parsed in your browser and shown in the same viewer engine the Studio uses, with the same render settings and the same Material panel for inspection. Nothing is sent to the server, so it is a private way to check a file before you upload it; when you are happy, upload it in the Model Editor to save it and publish a page.
/model-optimizer opens the same Optimize panel as Model Info in the Studio, without signing in and without uploading: drop a .GLB (max 150 MB) to see it at Original, Light, Balanced and Strong detail with real triangle counts, choose WebP, PNG or original textures (1K or 2K) and Draco geometry compression, and see the measured final size before you download it. Nothing is sent to the server.
A freshly uploaded model goes live on its own, no Save needed for the first publish. Every model is automatically paired with its own 3D-AR page (1 model = 1 page), so the moment the upload finishes, both the model and its page are already live. Any edits you make afterwards follow the normal Save flow with the "Unsaved changes" card.
Drop or pick a single .GLB and it renders instantly in a local preview, no upload yet. A "Preview" bar shows the file name and size so you can fine-tune Scene and Material first, then Confirm & Upload to save it (with your edits and any textures you added during the preview) or Discard to cancel. The preview even survives a page refresh, so nothing is lost by accident. Selecting multiple files at once skips preview and uploads as a batch.
The camera button in the viewer top bar (left of Download Model) captures a 4K (3840 x 2160) PNG render of the current model in its current scene and background. When it finishes, download it as ModelName_4K.png.
All your models appear as thumbnail cards below the viewer. Click a card to load that model; the active card is highlighted, and if you have unsaved changes a dialog asks you to save or discard first. Each card shows a source icon in front of its name (a wand for AI Generated, a cloud for Uploaded; BASE models carry no icon), a NEW badge on models you have not opened yet, and two corner buttons: info (opens More Details) and a three-dot Actions menu. Thumbnails update live as you edit color or scene, and cards briefly lock while a save is running.
Thumbnails are generated automatically when a model loads in the viewer. The system captures the current scene with all lighting and material settings applied. Thumbnails update whenever you modify the scene.
Find models instantly with live search. The Sort & Filter popup orders cards by Recently, Oldest, Largest, Smallest, Name A-Z or Name Z-A, plus a Reorder mode to drag cards into a custom order. A source switch filters cards by All, AI Gen, Uploaded, or BASE. Each card shows its thumbnail, name, size and source badge.
Swap the 3D file behind a model without losing anything else. The Change Model chip in the More Details popup replaces the model with a new .GLB (max 50 MB) while its page, public link, QR code and analytics are all preserved. Scene and material overrides reset so the new file starts clean. Available on models you uploaded yourself, not on BASE defaults or AI generations.
Every model card has a three-dot button that opens the full Actions menu. Press Esc or click outside to close it. Every item:
The info button on a card (or More Details in the Actions menu) opens the model hub. For every model, BASE library models included, it has two tabs under the header: Overview (everything below) and Optimize (the live file optimizer, detailed in the Model Info docs). The selected tab is shown bright and bold, and a small dot on the Optimize tab means a selection is waiting to be applied.
On phones the Model Editor keeps the canvas at the top of the screen, with the three editor tabs (Scene, Controls, Material) in a fixed half-screen bottom sheet below it. Your model cards live in a separate full-screen drawer: open it with the cubes button in the top bar, browse the three-column grid, and close it with the X, the arrow strip at the bottom, or simply by picking a model. A single Reset button at the top right of the canvas resets whichever tab is currently active, with the same confirmation popups as on desktop. Published 3D-AR pages and embeds are always fully mobile-ready, independent of the studio.
Master the lighting, environment, and camera behaviors to create the perfect 3D shot.
Each slider in the Scene and Controls tabs (and in the Page Designer right panel) has two controls on its right:
Typed values are automatically clamped to the slider's range and snapped to its step, so an out-of-range number is impossible. Model Orientation and Model Scale keep their own dedicated reset buttons, and Material rows use their own per-property resets, which return the value to the model's original GLB state rather than a Studio default.
Each of the three editor tabs has its own Reset Settings button at the very top. It resets only that tab, after a confirmation popup:
There is no cross-tab reset. On mobile, a single Reset button above the canvas resets whichever tab is active.
Customize the viewer backdrop:
Changes apply instantly and are mirrored live in your Page Designer mockup.
One-click looks for the whole scene. Presets belong to your account, not to a model: save a look while editing one model and apply it to any other. Seven built-ins are always available and cannot be deleted:
Click Add to save your current setup as a custom preset (names up to 30 characters, up to 20 custom presets per account). Applying a preset animates smoothly to the new values, the active chip shows a check, and a modified badge appears if you tweak settings afterwards. Custom presets are deleted from the X on their chip, with confirmation.
Controls the global illumination, reflections, and camera sensitivity.
Light Intensity (0 to 20): Brightness of the main directional light.
Ambient Strength (0 to 1.5): Fill light that brightens shadow areas. Increase if shadows appear too dark.
Azimuth (−180° to +180°): Rotates the sun horizontally around the model.
Elevation (−45° to +90°): Height of the sun. Lower values create longer, more dramatic shadows.
Controls the Shadow Radius (0 to 50).
0 = Sharp, hard-edged shadows.
50 = Very soft, diffused shadows.
Uses VSM soft shadows at 2048×2048 resolution (automatically 1024 on low-end devices) for consistent quality on the published page.
A fill-light slider (0 to 1.5) that lifts the brightness of shadow areas without moving the main light. Raise it if shadows look too dark, lower it for more contrast.
Automatically spins the model when idle.
Play Animation: Automatically plays embedded GLB animations. If the model contains multiple animation clips, select the active one from the dropdown.
Model Shadows: Master toggle to enable/disable cast shadows. Disabling can improve performance on low-end devices or when shadows are not needed.
Switch between five viewport display modes using the toggle buttons at the top-left corner of the 3D viewer. Render sits in the middle because it is the default, and each side strips something away: to the left the lighting goes, to the right the surface goes. Inspired by professional 3D tools like Blender, Marmoset and Sketchfab.
The default display mode. Shows the full PBR (Physically Based Rendering) output with textures, lighting, reflections, HDRI environment, and shadows. This is how your model will appear on the published page.
Displays the model with a flat grey material, no textures, no reflections. Useful for evaluating the model's shape, proportions, and overall form without visual distractions from materials.
Paints a checkerboard through the model’s own UV layout, so you can read how the texture is mapped. Squares that stretch into rectangles mean the UVs are distorted there; squares that change size from one part to another mean the texture detail is not evenly distributed; a visible break in the pattern is a seam. A mesh with no UVs at all is drawn in solid orange, so you can tell “no mapping” apart from “flat mapping”.
Draws each surface direction as a colour. It is the fastest way to spot flipped faces, broken normals and bad smoothing: a correct model reads as smooth colour gradients, while a problem area shows a hard seam or a patch that does not match its neighbours.
Shows only the mesh edges of the model, revealing polygon topology and edge flow. Adapts to your theme, lighter in dark mode, darker in light mode for optimal visibility.
Fine-tune viewer behavior, animation playback, camera constraints, and model positioning for the perfect interactive experience.
Control embedded 3D animations from GLB/GLTF files (the Animation section sits at the bottom of the Controls tab):
The play toggle can be stored in Scene Presets; the selected clip always stays with the model.
Create a smooth turntable effect for product showcases:
Manage user zoom interactions and define zoom boundaries per model:
Allow visitors to shift the model within the canvas:
Master control for real-time shadows:
The Reset Settings button at the top of the Controls tab resets only this tab, after a confirmation:
The Scene and Material tabs have their own Reset Settings buttons with their own scope; nothing crosses tabs.
Precisely position your 3D model within the viewer canvas without modifying the original file.
Re-orient the model on three axes without editing the original file, ideal for fixing a GLB that was exported facing the wrong way. Each axis has a slider, a numeric degree input, and a reset, plus a Reset All Axes button. Orientation is saved per model and never part of Scene Presets.
Restrict camera rotation to prevent viewing the model from behind:
Prevent users from rotating the camera below the model:
While Lock Back View or Lock Bottom View is enabled, Auto-Rotate is paused automatically, otherwise the turntable would spin straight into the blocked angle. Turn the lock off and rotation resumes.
Define the exact camera angle and zoom level that users see when the viewer first loads. While you frame the shot, the editor enters a focused mode: auto-rotate pauses and the panels step aside until you confirm or cancel.
The last section of the Controls tab sets how big the model really is. This is what AR uses when it places the object in the room, so getting it right here is what makes a chair land chair-sized on the floor instead of doll-sized. Changing the scale marks the model for a fresh AR export on your next Save, so the published AR file always matches.
A professional PBR (Physically Based Rendering) material system. Fine-tune every surface property directly in the browser with real-time feedback.
Identify and manage materials within complex scenes effortlessly.
Defines the surface's diffuse color (Albedo).
• Factor: Solid HEX color.
• Texture: Upload standard RGB image.
Controls light scattering.
0.0 = Polished Mirror.
1.0 = Matte/Rough.
*Map uses Green (G) Channel.
Defines material conductivity.
0.0 = Dielectric (Plastic/Wood).
1.0 = Metallic (Gold/Steel).
*Map uses Blue (B) Channel.
Adds fake geometric detail and depth to the surface without increasing polygon count.
Simulates soft shadows in crevices and corners for added realism.
Makes the material emit light, ideal for screens, neon signs, LEDs, and indicator lights.
Controls transparency, culling, and blending behavior.
Each material property row has its own reset button. It returns that property to the value stored in the original GLB file, not to a generic Studio default, so a reset always means "back to how the model shipped". Removing a texture asks for a confirmation first. To revert every material at once, use Reset Settings at the top of the tab.
Any material change (including renames and texture uploads) raises the "Unsaved changes" card with Save and Discard. Material edits are stored with the model and are never part of Scene Presets, so applying a preset can never overwrite your materials. Discard returns everything to the last saved state.
Upload textures per channel (Base Color, Roughness, Normal, etc.) in PNG, JPG or WebP format, up to 15 MB per file. Each channel has an upload and remove button for individual control. Textures are processed with correct color space handling: sRGB for color maps and Linear for data maps (roughness, metalness, normal).
Any texture whose longest side exceeds 2048 px is automatically downscaled on the server. The format and alpha channel are preserved, lossless WebP stays lossless, and images are never upscaled. The previous texture is kept until you Save, so Discard can always restore it. During an upload preview, textures stay local and are sent together with Confirm & Upload.
Note: AVIF and HEIC are accepted as Image to 3D inputs in the AI Generator, not as material textures.
Configure the visual interface, responsive layout, and Augmented Reality call-to-actions for your 3D-AR pages.
Design controls live in the right panel, matching the Model Editor. The Page Theme strip and sharing tools sit directly under the mockup.
Design with confidence using real-time synchronized previews for both screen sizes. Each frame renders its own device's saved values, so a mobile-only change appears only in the phone frame.
Capture your entire page design configuration as a reusable theme. Apply it to any page with one click, perfect for maintaining brand consistency across your product catalog.
Eight built-in themes are always available, Minimal, Neon, Elegant, Retro, Glass, Brutal, Azure and Cherry, alongside any you save yourself. All of them appear as horizontally scrollable chips in the theme strip. Click a chip to apply it to the current page. The active theme shows a green check, and a dot appears if you modify settings after applying. Built-in themes cannot be deleted or changed.
When you modify settings after applying one of your own themes, a dot icon appears on its chip: click it to quick-update that theme with the current look. Built-in themes cannot be overwritten, so a quick update on a built-in saves a new theme instead. Up to 20 saved themes per account.
Give each theme a descriptive name (up to 50 characters), e.g., "Minimalist White", "Brand Dark", "Holiday Theme". Duplicate names are prevented.
Hover over any of your own theme chips to reveal the delete button. A confirmation popup ensures you don't accidentally remove themes.
Themes are account-level: create once, apply to any page. Switch between pages and apply the same design configuration instantly for a unified brand experience.
Every model automatically gets its own page when uploaded or generated. The page title is auto-generated (e.g., "Chair 3D-AR") and can be customized anytime. There is no separate page creation, each model is always paired 1:1 with its page.
You can set a page Live or Offline from the Library (status badge) or from a model card's three-dot menu or its More Details popup (Set Offline / Set Live). In the three-dot menu the eye icon turns red while a page is Offline, so the state is visible at a glance. Offline pages stay saved and can be brought back live anytime, without deleting them.
Everything under the AR Button tab, section by section.
Fully customize the "View in AR" button to match your brand guidelines.
On the Free plan the Bottom Right corner is reserved for the "Powered by BASE" badge: neither the AR button nor the logo can use it. If a theme or an older saved value places them there, Studio visibly corrects it (AR button to Bottom Center, logo to Top Left) on both devices and opens the save card so nothing changes silently. AR button, logo and viewer-button settings are saved separately for Desktop and Mobile.
You can optionally upload a custom .usdz file for iOS AR Quick Look from the AR Button tab, under Visibility. It is optional, iOS AR works without it. Only .usdz files up to 50MB are accepted. The card works like a toggle: upload to set it (the label turns green with a checkmark), click again to remove the file. The .USDZ file is shared by both devices.
When desktop visitors click the AR button, a QR popup appears to bridge them to mobile AR. You can fully customize the popup's title and description text for each page, perfect for brand messaging, multilingual content, or product-specific instructions. Because this popup only ever appears on desktop, its settings are intentionally shared rather than per device.
Replace the default "View in AR" heading with your own text, up to 100 characters. Use it for brand names, product titles, or localized messaging (e.g., "AR ile Dene", "Voir en RA").
Edit the instructional text below the QR code, up to 300 characters. Guide visitors on how to use AR, add product details, or write a call-to-action tailored to your audience.
The editor modal displays an exact replica of the real QR popup, same styling, same layout, with a live mini preview built around your page's real QR code. What you see in the editor is exactly what your visitors will see on desktop.
Each page stores its own QR popup text independently. Customize different messages for different products or models. Changes appear instantly in the preview and go to the live public page when you save.
Go beyond text, fully customize the QR popup's visual appearance with 4 independent color pickers. Match your brand palette or create themed popup designs that stand out.
#FFFFFF
#000000
#000000
#000000
Color changes are stored per-page and go to the live public page when you save. Edit colors in the QR Popup Editor modal alongside the text content.
The logo controls live in the lower half of the Toolbars - Logo tab and share that tab's device switch.
Upload your logo (PNG, JPG, JPEG or WebP, up to 5MB) and position it with precision.
Create distinct layouts for different screen sizes, or keep them in lockstep.
What visitors get on the live page, and how those buttons look.
Choose which buttons appear for visitors on the live page, and set a shared color theme for the top-right cluster.
A live diagnostics HUD your visitors can open on the published 3D / AR page. It is off by default on every page, and you turn it on per page from Page Designer → right panel → Toolbars - Logo tab → Viewer buttons → Performance.
Turning it on or off is realtime: visitors who already have the page open see the change without reloading. Best used for client demos, device testing and support tickets. Leave it off for a fully clean, branded presentation.
Auto-generated QR code linked to your live AR page. Download as image for print materials, packaging, or in-store displays.
A clean, shareable URL hosted on a global CDN. Copy to clipboard with one click. No coding or setup required.
Ready-to-use <iframe> snippet. Paste it into any website to display the interactive 3D viewer with full AR support.
The share strip under the mockup gives you four one-click actions for the selected page. Until the page has a link, the link-based buttons stay disabled.
The public REST API is live at https://studio.basebros.com/api/v2. Authenticate with an API key (Authorization: Bearer bpk_live_...); start generation jobs, download assets, update page settings, publish or unpublish pages, and receive signed webhooks when jobs finish. The API is server-to-server only and shares your account's generation quota. API access is enabled per organization and keys are issued by BASE, reach out via Contact to get started. For in-page integration, the iframe Embed Code remains the supported path - see the Web & Deploy section for details on both.
Page Designer edits are a live preview until you save them. Page Designer and Model Editor share one save card in the bottom-right corner: it reads Unsaved changes with a second line naming the scope - Model, Page or Model + Page - plus Save and Discard. One click saves everything, and it is live on your public page the moment it finishes. There is no separate publishing step. Saving always runs in a fixed order: the model is published first, then the page design. If the page step fails you are told the model is live and offered a retry. Saving never flips a page's Live/Offline state, an Offline page stays Offline. Discard reverts both the model and the page to the last saved state after a single confirmation.
On phones the Page Designer keeps the desktop and mobile mockups at the top of the screen with the settings tabs (AR Button, Toolbars - Logo) in a fixed half-screen bottom sheet. The share strip sits directly under the mockups with four compact actions: Info, Copy, Share and Open. Your page themes appear as a single-row theme strip between the share strip and the sheet: swipe horizontally through the chips and use the round + button pinned on the left to save the current design as a new theme. All device-specific settings, the sync checkbox and theme saving behave exactly as on desktop.
How the Augmented Reality experience works across devices, from QR-based desktop flow to native mobile AR.
The AR experience adapts to the visitor's device automatically:
The QR popup is a branded overlay that appears on desktop when a visitor clicks the AR button. It displays a scannable QR code linking to the AR experience, along with the AR icon and a close button. The title, description text and all four colors (card background, title, text, close button) are fully customizable per page via the QR Popup Editor in Page Designer.
The AR button is fully customizable per device (desktop and mobile independently, or kept identical with the Use the same on desktop and mobile checkbox). Use the Show AR button toggle under Visibility in the AR Button tab to enable or disable it on your live page. When disabled, the AR button won't appear to visitors and the other AR controls dim. Control the text, colors, font size, opacity, corner radius, and position. All settings are configured in the Page Designer section.
AR requires an HTTPS connection (mandatory for camera access). The embedding page must include the allow="camera; xr-spatial-tracking; accelerometer; gyroscope; magnetometer; fullscreen; autoplay" permissions in the iframe tag. See the Web & Deploy section for the complete embed snippet.
Technical guide to embedding 3D/AR experiences with performance and security best practices.
<iframe
src="https://studio.basebros.com/view?id=YOUR_PAGE_ID"
width="100%"
height="600"
style="border:0;"
loading="lazy"
allowfullscreen
allow="camera; xr-spatial-tracking; accelerometer; gyroscope; magnetometer; fullscreen; autoplay">
</iframe>The allow attribute is critical for AR functionality. Without these specific tokens, browsers will block sensor access:
Augmented Reality features will not work on non-secure (HTTP) domains due to modern browser security policies.
Avoid fixed pixel heights. Use modern CSS aspect-ratio to maintain the perfect frame on all devices.
.ar-container {
width: 100%;
aspect-ratio: 4 / 3;
border-radius: 12px;
overflow: hidden;
}
iframe { width: 100%; height: 100%; }
Always add loading="lazy" to your iframe tag. This ensures the 3D engine only initializes when the user scrolls it into view, boosting your site's LCP score.
Include allowfullscreen so users can expand the 3D viewer to monitor size for detailed inspection.
# 1. Is the Studio up? (no auth required)
curl -s "https://studio.basebros.com/api/v2/status" \
-H 'BASE-Version: 2026-08-09'
# 2. Does my key work?
curl -s "https://studio.basebros.com/api/v2/organization" \
-H "Authorization: Bearer $BASE_API_KEY" \
-H 'BASE-Version: 2026-08-09'API access is enabled per organization, reach out via Contact to activate it. Keys are issued by BASE and shown once: bpk_live_ for production and bpk_test_ for simulation (test jobs never trigger real generation or spend quota). Keep your key in your backend's secret manager. The API is server-to-server only: no CORS headers are ever sent on /api/v2, so a key can never run in a browser. This is deliberate, it physically prevents keys from leaking into client code.
Over REST you can start 3D generation jobs, poll their status, download the resulting assets, update page presentation settings, and publish or unpublish pages. Webhook deliveries notify your backend when a job finishes and are signed per the Standard Webhooks spec (webhook-id, webhook-timestamp, webhook-signature). API and dashboard share the same monthly generation quota. A machine-readable definition is available at GET /api/v2/openapi.
Comprehensive guide to understanding your real-time 3D-AR analytics dashboard, Quick Stats, collapsible cards, charts, and data-driven insights.
The Analytics dashboard provides real-time visibility into how visitors interact with your 3D-AR pages. The dashboard features 12 Quick Stats metrics, 8 collapsible card sections with Live indicators, and auto-refreshes every 60 seconds. Use the Page Selector dropdown to switch between All Pages overview and individual page analytics. Filter data by All Time, 30D, 7D, 24H, or Custom date range pills. Every card section is collapsible, click the header to expand or collapse. Cards always open expanded when you load the dashboard.
The dashboard is organized into 8 collapsible card sections (Quick Stats, AI Generation, Traffic, Leaderboard, Distribution, Sources & Activity, Visitor Insights, Engagement & Duration). Each card has a green icon header, a pulsing Live indicator badge, and a chevron toggle. Click any card header to collapse or expand its content. Cards always start fully expanded on load.
The total number of page loads across all your published 3D-AR pages within the selected time range. Each time a visitor opens your page, it counts as one view. Animated counter displays the value with a smooth GSAP transition.
The number of distinct browsing sessions. A session groups all views from the same visitor within a continuous browsing window. Multiple page views by the same visitor in a single session count as one unique session.
The number of currently published 3D-AR pages in your account. This count reflects live pages only, deleted pages are excluded. Use the page selector dropdown to view analytics for specific pages.
The count of 3D models you have uploaded to the Studio. Default/sample models are not included in this count. This number reflects your actual model usage against your plan quota.
The number of times visitors launched Augmented Reality mode. Each tap of the "View in AR" button counts as one activation. Requires a compatible device (iOS with ARKit or Android with ARCore). High AR rates indicate strong product interest.
Total discrete interaction events with 3D models, including rotation gestures, zoom actions, and tap events. A high interaction count relative to views indicates visitors are actively exploring your 3D content rather than passively viewing it.
The number of page views originating from QR code scans. When visitors scan a QR code generated from the Page Designer, the visit is tagged with a QR source. Use this to measure the effectiveness of your print materials, packaging, and in-store displays.
The average time visitors spend actively interacting with your 3D models per view. Tracks actual hands-on engagement (rotating, zooming, inspecting), not passive idle time. Displayed in human-readable format (e.g., 1m 23s).
The percentage of sessions with only one page view. A lower bounce rate is generally better. Industry average for product pages is 40-60%. 3D-AR pages often have unique patterns since a single page can provide a complete experience.
The number of 3D models generated using AI by describing them with a text prompt. Each successful Text-to-3D generation is counted. Available with the four text-capable engines (Meshy, Tripo, Rodin, Hunyuan); TRELLIS and Pixal3D are image-only. Tracks your AI generation productivity across all sessions.
The number of 3D models generated using AI by uploading a reference image. Each successful Image-to-3D generation is counted. Available with all six AI engines (Meshy, Tripo, Rodin, Hunyuan, TRELLIS, Pixal3D). Tracks your visual AI generation usage.
The cumulative total time visitors have spent actively interacting with your 3D models. Unlike Avg View Duration (per-view average), this shows the total engagement time across all views. Displayed as formatted duration (e.g., 5h 30m).
The Traffic card contains two side-by-side time-series charts. Views (green, toggleable between Line and Bar) and Sessions (blue, line chart) show daily traffic patterns over the selected range. Located inside the collapsible Traffic card section with Live indicator.
Four horizontal bar charts inside the collapsible Distribution card: OS (purple), Devices (purple, Mobile/Tablet/Desktop), Browsers (orange), and Locations (green, top countries). All charts respond to page selector and date range filters.
Shows where your traffic originates from (blue horizontal bar chart). Direct means visitors typed your URL or used a bookmark. Other entries show referring domains. Located in the collapsible Sources & Activity card alongside the Activity Heatmap.
A 7×24 grid with green-toned cells showing visitor activity by day of week and hour of day. Darker green cells indicate higher traffic. Use this to identify peak engagement windows for scheduling campaigns. Hover over any cell to see the exact view count.
The Leaderboard card shows your top-performing pages ranked by views and sessions. Only visible when All Pages is selected in the page selector (hidden in single-page view). Useful for identifying your best content and understanding relative page performance.
The entire dashboard auto-refreshes every 60 seconds to show the latest data. The pulsing green Live indicator badge on each card header confirms real-time data flow. You can also manually refresh anytime using the Refresh icon button in the header.
A three-panel card providing deeper visitor analysis. Shows who visits, where they come from, and when they're most active.
14:00 - 15:00), period label (Morning/Afternoon/Evening/Night), and view count.
A visual 4-step engagement funnel showing how visitors progress through increasingly deeper interactions: Page Views → 3D Interactions → QR Popup Opens → AR Activations. Each step shows count, percentage of previous step, and a progress bar. Summary stats include drop-off %, end-to-end conversion %, 3D rate %, and QR rate %.
A doughnut chart showing how long visitors stay on your pages, categorized into 4 time buckets: Quick (<30s, red), Short (30s-2m, amber), Medium (2m-5m, green), and Long (5m+, blue). Includes a custom HTML legend with percentages, plus summary stats for average duration and bounce rate.
Best practices and expert recommendations for getting the most out of the BASE 3D-AR Studio.
The right lighting setup makes or breaks the visual quality. Here are proven starting points:
Create 3D models from text prompts or images using AI-powered generation
The 3D Generator uses six AI engines, Meshy, Tripo, Rodin, Hunyuan, TRELLIS and Pixal3D, to create 3D assets from text descriptions or reference images (TRELLIS and Pixal3D work from images only). Meshy is the default engine and Image to 3D the default mode. No 3D modeling skills are required: describe what you want or upload an image, and the engine handles the rest. You can run up to 3 generations at the same time and queue 10 more. Every result is a GLB file, with PBR materials by default.
Describe your 3D model in natural language, up to 500 characters with a live counter under the prompt box (at least 10 characters to generate). The selected engine interprets your prompt and generates a matching 3D asset; Text to 3D works with Meshy, Tripo, Rodin and Hunyuan. Works best with clear, descriptive prompts: "A vintage wooden chair with curved armrests" or "A futuristic robot with metallic armor". An estimated time badge in the prompt box updates as you change the engine and its options.
Upload reference images (PNG, JPG, WEBP, AVIF, HEIC up to 20MB each) and the AI reconstructs them as a 3D model. How many images you can attach depends on the engine: up to 5 with Rodin, up to 4 with Meshy, Tripo and Hunyuan Pro, and a single image with Hunyuan Rapid, TRELLIS and Pixal3D. The counter under the dropzone always shows the current limit, and switching engines keeps your main image while trimming any extras over the new limit. AVIF and HEIC files are converted automatically before generation. Best results with clear, well-lit images showing the full object.
Choose from six AI engines, listed in the picker in this order: Meshy, Tripo, Rodin, Hunyuan, TRELLIS and Pixal3D. Meshy is selected by default. Each row in the picker shows a live status dot, a short description, the input modes the engine accepts (a text and an image glyph; an unsupported mode is dimmed) and two ratings, Speed and Detail, from one to three dots. The ratings compare engines with each other; the Version picker uses the same two columns to compare the versions of a single engine. TRELLIS and Pixal3D work from images only, so selecting either one switches the panel to Image to 3D. Each engine has different strengths:
Polygons are chosen as named levels, not raw numbers, and each button shows the level name only. Every engine offers its own ladder, ordered from lightest to heaviest: Meshy Low · Medium · High · Max, Tripo Low · High, Rodin Low · Medium · Ultra · Original, Hunyuan Low · Medium · High, TRELLIS and Pixal3D Low · Medium · High · Max. The top step is the default on Tripo, Rodin and Hunyuan. On Meshy, TRELLIS and Pixal3D the default is High, one step below the top; Max on those engines produces very large files. A lower level makes the file smaller; it never adds detail. When you switch engines, your level is kept if the new engine has a step with the same name; otherwise the new engine's default applies. The names are shared, the mechanisms are not:
Meshy is the default engine, described in the picker as top textures, most options and rated 2 of 3 for Speed and 3 of 3 for Detail. It supports Text to 3D and Image to 3D from one image or up to 4 images for multi-view. In Text to 3D it works in two stages, a geometry preview followed by a texture refine pass. Pick Version 6 or 7 next to the engine; Version 7 is the default. The feature rows follow in this order: Polygons, Ultra Mode, Texture, Texture Quality and Topology.
Tripo, described in the picker as fast and reliable, is the only engine rated 3 of 3 for Speed (Detail 2 of 3), which makes it the natural pick for rapid ideation and prototyping. It supports Text to 3D and Image to 3D from one image or up to 4 images. Pick v3.0 or v3.1 next to the engine; the feature rows follow in this order: Polygons, Ultra Mode, Texture and Texture Quality. For a game-ready low-poly asset set Polygons to Low, which hands the job to Tripo’s Smart Mesh line.
Rodin, described in the picker as sharp edges, clean mesh, brings sculpt-level geometric detail and fine control over quality, texture and topology. It runs through fal.ai and is rated 2 of 3 for Speed and 3 of 3 for Detail. It supports Text to 3D and Image to 3D from one image or up to 5 images, the most of any engine. Pick the generation (Gen 2, Fast or Gen 2.5) next to the engine: Gen 2.5 is the default and the most detailed, while Fast is the quickest and tops out at the lower tiers and face counts.
Hunyuan, described in the picker as best for simple shapes and rated 2 of 3 for both Speed and Detail, ships two different pipelines behind a single name. The Engine row, placed below Polygons, decides which one you get and changes which other options are available. Pro is the full-control path: pick the model version, choose a polygon level from Low (30K) to High (the native mesh of about 500K), and feed it up to 4 reference images for multi-view (the first is the front view, the next three become back, left and right). Rapid drops those controls in exchange for speed. Both support Text to 3D and Image to 3D, and both can produce full PBR materials.
TRELLIS, described in the picker as quick with real materials and rated 2 of 3 for both Speed and Detail, turns a single image into a 3D model with full PBR materials. It does not accept text prompts, and it takes one image because fal’s multi-image endpoint averages the views instead of reconstructing them; for multi-view pick Meshy, Tripo, Rodin or Hunyuan Pro. There is no Version control, since fal serves a single TRELLIS 2 model; all three controls sit in the feature rows, in this order: Quality, Polygons and Texture Quality.
Pixal3D, described in the picker as solid all round detail and rated 1 of 3 for Speed and 2 of 3 for Detail, turns a single image into a 3D model with full PBR materials. Its input pipeline follows the TRELLIS family, so its controls mirror TRELLIS: Quality, Polygons and Texture Quality, with no Version control. It does not accept text prompts or multiple images.
From idea to 3D model in five simple steps:
You do not have to wait for one model to finish before starting the next: up to 3 generations run at the same time, on the same engine or across different ones. Start a fourth and it does not get refused: it joins your queue, and up to 10 can wait there. The moment one of your three finishes, fails or is cancelled, the next one in line starts on its own. Both limits are the same on every plan, Free included. A job selector at the top of the production panel keeps every job one click away: it shows a pulsing status dot, the active job's name and a "N of M generating" counter, and opens a list of your jobs.
Options live in the right-side production panel as open switcher rows: the section name sits on the left and its choices sit side by side underneath, so every setting is visible at once with no dropdown to open. The panel starts with the job selector (visible while jobs are running) and the Image to 3D / Text to 3D tabs (Image to 3D is open by default), then the scrolling form: Prompt, then AI Engine (the engine and version pickers, the only two controls that open a popup, and they open downward; TRELLIS and Pixal3D have no version picker), then the feature rows, which differ per engine (see the engine cards above). The green Generate 3D Model button sits in a fixed action box at the bottom of the panel. Settings apply to both Text-to-3D and Image-to-3D, and the live ETA updates on every change in the bottom-left corner of the prompt box.
You can cancel an ongoing generation at any time with Cancel generation, the borderless button with a red X at the bottom of the generating card, or from the job's row in the job selector. The generating screen shows a continuously spinning ring and an elapsed timer rather than a percentage bar, because the finish time is not known in advance. Important: It still counts as a new model this month, since the AI provider has already started processing. It does not take a live spot. If a job fails on the provider's side or times out, it doesn't count.
When generation completes the right panel shows the Your model is ready card, subtitle "Successfully generated and saved to your library." At the top is a small live 3D preview of the new model that you can rotate and zoom. Under the title the generation time is printed as a quiet line in quotes with a green stopwatch icon. Then your prompt and the configuration badges, and finally Download and Live page side by side. There is no See the model button: the model is already in the slider below the stage, so clicking its card opens it. To start over, the bottom action button becomes + New Generation. The model is already saved to your library at this point.
The This month meter counts new models. Every new model counts the same: Text-to-3D, Image-to-3D, a .GLB upload and Change Model each use 1. Free: 3 a month. Pro: up to your capacity each billing month. The allowance resets on your billing date (on Free, the day of the month you joined). When 2 are left, a warning appears before you generate; at zero, a limit window shows when it resets and how to get more (Go Pro, Increase capacity or Contact us). Deleting or archiving a model does not give one back. Generations that fail on the provider side don't count.
Your plan sets how many models can be live at once: Free: 3, Pro: your capacity, Enterprise: custom. Uploads and AI generations share this one pool. Archiving or deleting a model frees a live spot. The BASE sample models that come with your account don't count, and archived models don't either. Downloading a GLB is always free. A compact badge in the bottom-right corner of the 3D canvas shows Live models and This month; jobs still in flight appear as a reserved suffix, for example 38/40 (+2). The badge turns yellow above 60% usage and red when it's full.
Every engine card in the AI Engine list carries a live status dot that refreshes about every 30 seconds: ready, busy or offline. An offline engine is dimmed and cannot be selected; if it was your current choice, the generator switches to the first available engine automatically. Each engine also has its own capacity pool: if the pool is momentarily full when you press Generate, Studio tells you and suggests available alternatives that you can switch to with one click, your prompt and images are kept.
Every generated model can be downloaded as a .GLB file, the industry standard for web 3D, or as a USDZ for iPhone and iPad, free and without touching any quota. The Download popup lets you choose the file: a triangle level (Original / Light / Balanced / Strong), WebP, PNG or original textures at 1K or 2K, Draco compression on or off, and whether your saved material edits are baked in. The summary shows the real file size (a custom mix is measured in the background first) and one button downloads exactly that file. Use it anywhere: import into Blender, upload to other platforms, or embed on your website.
Generation time depends on the engine, its options and the provider’s queue at that moment, so no fixed figure applies. The Speed dots in the engine picker compare the engines (Tripo is the only one rated 3 of 3, Pixal3D the only one rated 1), and the live ETA in the bottom-left corner of the prompt box estimates your exact settings: it is calibrated on completed jobs and also learns from your own finished generations on this device. Higher polygon levels, Ultra Mode, 4K textures, Quad topology (Meshy and Rodin) and Tripo’s Detailed texture quality add time; Draft quality, lower polygon levels and Texture None shorten it. The generating screen shows a spinning ring with an elapsed timer (there is no percentage bar).
When generation completes, the model is automatically saved to your library, no extra click needed, and its own 3D-AR page is created at the same time. It becomes available in the Model Editor and Library with full editing capabilities, scene, materials, page design, and AR deployment.
On phones the 3D Generator mirrors the Model Editor layout: the 3D preview canvas stays at the top and the generation form lives in a fixed half-screen bottom sheet, with the Image to 3D / Text to 3D switcher (Image to 3D open by default) at the top of the sheet and your quota badge in a single centered row right below it. The Generate 3D Model button is pinned as a fixed bar above the bottom navigation so it is always reachable while you scroll the options. Engine choice, generation options, multi-generation and the job selector all work exactly as on desktop.
Stay informed about the latest Studio updates, new features, improvements, and system status.
A dedicated full-page view to see all Studio updates. Access it from the Account Popover at the bottom of the sidebar, click your profile, then click Announcements.
A collapsible real-time health dashboard at the top of the Announcements page. Shows live status for 6 core systems with animated pulse indicators:
Green pulse = operational. The card can be collapsed to save space.
Real-time UTC clock displayed in your Account section.
A live real-time UTC clock displayed in the Account > Membership Card section. Updates every second and shows the current time in HH:MM:SS UTC format.
Detailed technical statistics, live file optimization, downloads and management actions for each 3D model.
Click the info icon (ℹ) or the three-dot menu on any model card in the Model Editor to open the More Details popup. It is the hub for that model and its 3D-AR page together. For every model, BASE library models included, it has two tabs under the header: Overview and Optimize (described below); The Overview shows the model thumbnail, name, file size, upload date, and a detailed statistics grid with 8 technical metrics:
The popup also shows the linked 3D-AR page: its QR code (with a download button), page name with inline rename, Live/Offline badge, the public link with a copy button, and shortcuts to Analytics and Share Page.
Beyond statistics, the More Details popup lets you manage the model and its page from one place. The same actions are also available directly on the card through the three-dot menu:
The Optimize tab of the More Details popup works directly on the GLB that your 3D-AR page serves. It has three parts, top to bottom: the level cards, the Textures and Geometry rows, and the Final file summary.
Library models: the first Apply creates your own copy of the BASE model (the shared original is never changed) and your card, page and live link switch to that copy. Not available: models whose live file has no source stamp yet, and browsers without Web Workers; without OffscreenCanvas the Textures row is disabled. While birth compression is still running, Apply waits a moment. Errors are shown inside the panel and never change the live page.
Every Download button in the Studio (model card menu, More Details footer, Library, 3D Generator) opens the same Download model popup: a small model preview, the file name and a facts strip at the top, then six rows where you decide what the file contains. There are no presets to pick; the system builds exactly the file you describe and names it accordingly:
The summary under the rows shows the file size in real bytes: a ready file shows it at once, and any other mix is built in the background so its measured final size appears after a short shimmer (very large files, and large files on phones, show the size when the download finishes). Then come the triangle, geometry and texture facts and one button: Download GLB or Download USDZ. When your selection matches a ready file (the stored live file, the -original file, the PNG and uncompressed -full file, or the USDZ) that file is served as it is; any other mix is built in your browser on demand, gets a -light, -balanced, -strong or -custom suffix and stays yours only: it never changes the live page (use the Optimize tab for that). One download runs at a time, progress shows inside the popup and the X cancels. Downloads are always free and unlimited, and they never use any quota.
If a model looks wrong or misbehaves, open the card's three-dot menu and choose Report Model (available for your models and the default BASE models alike). Pick one of seven categories: Looks broken, Does not load, Wrong material, Missing texture, Wrong size, Slow or laggy, or Something else. Add an optional description (up to 2000 characters) and up to 4 screenshots (5MB each, iPhone HEIC accepted).
Manage your profile, see your plan and usage, and change your capacity.
The Account page shows your profile (username, email, profile picture), your plan with a live UTC clock and join date, and your two usage meters: Live models against your capacity and This month, with the date it resets. Open Plans from Go Pro or Upgrade Plan (Free) or Change capacity (Pro). The Subscription card holds Change payment method, Invoices & billing and Cancel subscription. In the tools row, Restore Defaults brings back only the BASE sample models you removed, with their 3D-AR pages; samples you kept or edited stay exactly as they are, and samples never count toward your meters. Export All downloads every model as one ZIP.
There is no separate Settings page in the Studio menu. Your account details (name, email, password, plan) live in Account Settings (account pill at the bottom left), and Dark mode is in the same account menu.
From the Profile card you can change your password and change your e-mail; both confirm with a verification code sent to your inbox. Changing your e-mail keeps everything: your models, pages, analytics and billing move to the new address (if a 3D generation is still running, wait for it to finish first). Delete account lives in the same card and is a two-step confirmation: BASE e-mails you a code, you enter it, and the account with all its models, pages and analytics is permanently removed. It cannot be undone, so download anything you want to keep first (Export All in the Account tools packs every model into one ZIP).
$0, no card needed. Up to 3 live models and 3 new models a month, AI-generated or your own .GLB. The full viewer, scene and material editing, AR and publishing are included. Free pages show the Powered by BASE watermark, and the AR button's Bottom Right position is reserved for it. The BASE sample models in your library don't count.
$0.50 per model, per month at our launch price, minimum $20 a month for 40 models. You choose your capacity, any number from 40 up to 2,000: it sets both your live models and your new models each month, AI-generated or uploaded, same price. No BASE watermark. Works in Studio and through the API (on request). Billed monthly, upfront.
For large catalogs and high monthly volumes: custom capacity and monthly volume, custom 3D projects built by our team, and dedicated support and onboarding, with pricing sized to your catalog. Need more than 2,000 models? Use Contact sales in Plans and the BASE team sets it up. Enterprise limits are managed by BASE: when you reach one, the limit window offers Contact us and Manage models.
Set Offline takes a model's public page offline, but the model stays in your library and still counts as live. Archive takes the model and its page offline, stops it counting as live and keeps everything (settings, materials, page) with no time limit; restore it from the model's menu or the Archived filter in the Library. Delete removes the model and its page for good, and its link and QR code stop working. Archiving or deleting frees a live spot, but neither gives back a model from This month.
A raise applies right away. You pay the prorated difference for the rest of the billing month and get the matching share of extra new models for this month at once; your full new capacity applies from your next billing date. Example: you're on 40 models ($20 a month) and raise to 60 ($30 a month) with 15 of your 30 days left. You pay about $5 now (half of the $10 difference) and get 10 extra new models this month. From your next billing date you pay $30 and get 60 new models a month. A difference under $0.50 isn't charged now; it's added to your next invoice. Polar emails the invoice.
A lower applies on your next billing date, with no charge now. Change scheduled confirms the date, the new capacity and price, and how many models you'll need to archive or delete: with 52 live models and a new capacity of 40, that's 12. Use Choose now to pick which models stay live (we apply it that day), or decide later. On the date, if you still have more live models than your plan includes, the Downgrade complete screen opens: select models and use Archive selected or Delete selected until you're within your plan, or Increase capacity instead. It stays open until you're done. If you don't choose within 15 days, we keep your 40 most recently added models live and archive the rest automatically. Nothing is ever deleted automatically.
Switch to Free in Plans, or Cancel subscription in Account, keeps Pro until the end of the month you paid for, and you can Resume Pro any time before then. After that, Free applies: 3 live models, 3 new models a month and the BASE watermark. It works like lowering: Change scheduled shows how many models you'll need to archive or delete, Choose now is optional, and on the day the Downgrade complete screen (Go Pro instead is there too) gives you 15 days before we keep your 3 most recently added models live.
If a renewal payment doesn't go through, new models pause (AI generations, uploads, Change Model and API jobs) while your live models keep working. A banner and the Payment failed window offer Update payment, and capacity changes wait until it's paid. Polar retries your card automatically for up to 21 days. If it's still unpaid then, your plan ends and you're on Free: the Your plan ended after a failed payment screen works like Downgrade complete, with the same 15 days. Archived models come back when you go Pro again.
Restoring an archived model needs a free live spot, and it doesn't use a model from This month. If every spot is in use, the Your Plan Is Full window opens: Go Pro (Free) or Increase capacity (Pro) to make room, or Archive another model to swap one out. Choose live models in the Library's Archived view does both in one step: it archives the models you untick, then restores the ones you tick.
Prices are in USD. In most countries VAT is already included; in the US, Canada and India, sales tax is added at checkout. Polar, our merchant of record, works it out and shows it before you pay, and the same rule applies to a prorated raise. Every charge has an invoice: open Invoices & billing in Account to see them, and Polar emails a receipt for each payment. Update your card with Change payment method.
Tailored 3D modeling and custom viewer experiences built exclusively for your brand by the BASE BROS team.
Custom Projects is an Enterprise plan exclusive service where the BASE BROS team designs and develops a fully tailored 3D viewer experience for your brand. This includes custom UI/UX design, bespoke viewer functionality, unique interaction flows, and branded visual elements - all built from the ground up to match your exact requirements.
Fully branded viewer interface tailored to your corporate identity. Custom layouts, color schemes, typography, and interactive elements designed specifically for your use case.
Bespoke viewer features built to your specifications. Product configurators, custom annotation systems, special interaction modes, data integrations, and any unique functionality your project demands.
Priority support, a dedicated project manager, custom onboarding, and support terms and SLA agreed per account for your mission-critical deployments.
Custom deployment options tailored to your existing infrastructure. API integrations, white-label solutions, and specialized embedding workflows to fit your platform perfectly.
Custom Projects is available exclusively for Enterprise plan subscribers. To discuss your project requirements and get a tailored proposal:
1. Reach out to [email protected] with a brief description of your project.
2. Our team will schedule a discovery call to understand your needs.
3. We deliver a detailed proposal with timeline and scope.
4. Development begins with continuous feedback and iteration.
Our in-house 3D team personally creates and edits production-ready models for your products.
Don’t have a 3D model? No problem. With the Custom 3D Modeling service (Enterprise plan exclusive), our in-house team of 3D artists works directly with you - one-on-one - to create or edit high-quality, web-optimized 3D models from your product photos, CAD files, sketches, or reference images. Every model is handcrafted and delivered ready to use with BASE Studio’s full AR and viewer capabilities.
Send us product photos, CAD drawings, or even rough sketches. Our 3D artists will create a detailed, accurate 3D model that faithfully represents your product in every angle and detail.
Every model is optimized for web performance - lightweight file sizes, clean topology, PBR materials, and AR-ready GLB format. Instant loading, stunning visuals.
PBR materials, realistic textures, and precise lighting setups ensure your 3D models look stunning in the viewer and AR. Perfect for e-commerce, product showcases, and marketing.
Iterative feedback process with revision rounds included. We work closely with you to ensure every detail is perfect, with fast turnaround times for your production schedule.
Custom 3D Modeling is available exclusively for Enterprise plan subscribers. To get your custom 3D model:
1. Email [email protected] with your product photos or reference materials.
2. Our team personally reviews your project and provides a timeline.
3. One-on-one modeling begins with progress previews at each milestone.
4. Final model is delivered directly to your BASE dashboard, ready to deploy.
Can't find what you're looking for? Our team is here to help.
Questions about plans, features, or partnerships.
[email protected]Bug reports, API questions, or integration help.
[email protected]BASE BROS Bilişim Hizmetleri A.Ş.
Girne Mah. Girne Cd. No: 125/2, Maltepe / Istanbul / 34852, Türkiye
Teknopark Istanbul, Sanayi Mahallesi, Teknopark Bulvarı, No:1/4C Blok, 3/307, Pendik / Istanbul / 34906, Türkiye
LinkedIn · X · Instagram (all @basebros3d) · basebros.com
[email protected]: general questions, plans, billing, sales and partnerships. [email protected]: technical support, bug reports, API and integration questions. We typically answer within 2 hours on business days.
Open Report a Bug from your account menu or from this page. Fill in a Subject, pick a Category (UI / Design, Performance, Upload, 3D Viewer, Editor, AI / Chat, Other) and describe what happened. Attach up to 4 screenshots (5MB each, drag & drop or click, iPhone HEIC accepted). The page you were on is attached automatically, so the team sees exactly where the problem occurred.
The My Reports tab lists every report with a status badge: Open, In Progress, Resolved or Closed. Team replies appear under each report, and an unread badge shows up in real time when a new reply arrives. Messages works the same way for general support conversations, and supports the same image attachments.
Your command center for every published page. See all your 3D-AR pages, their models and live links in one searchable place, then copy, open, analyze or export in a click.
Library gathers every page you own (Base models, AI-generated creations and your own uploads) into a single modern list. The header shows three live stats: Pages, Active and Total Views. No more hunting through tabs: grab a shareable link, preview the model, jump straight to a page's analytics, or download the whole catalog as a spreadsheet.
Every row shows its public view URL. • Copy the link to your clipboard. • Open the live page in a new tab. • Share & Embed: QR code, Copy Link, Copy Embed. • Download QR to print or package; scanning opens the live page where visitors tap View in AR.
A mini thumbnail and the model name for each page, with a clear type tag: • Base · AI-Gen · Upload
See total views per page, and tap the chart icon to jump straight into that page's full analytics dashboard.
Each row in table view carries everything about one page, left to right:
Gallery cards show the same information, with the Active badge and type tag placed on the thumbnail itself and the actions in a single row underneath.
Every row and gallery card ends with the same compact action set, in this order:
One model, one page: every model is paired with its 3D-AR page, so deleting a page also permanently deletes the 3D model behind it. The confirmation always states exactly what will be removed. Info and Download appear only on rows that have a model attached.
Find any page instantly and shape the list exactly how you like.
Switch between a compact table for scanning many pages and a visual gallery of model thumbnails. Your choice is remembered, and on mobile rows fold neatly into cards.
Download your catalog with one click from the Export menu, perfect for reports, backups or sharing with your team. Exports respect your current search, filter and sort, and include title, status, type, model, total views, created/updated dates and the live link.
Tap Select Models in the toolbar (just left of the filter chips) to enter selection mode. Model images become the selection target, everything else on the row dims and goes inert so nothing is triggered by accident. Tap any row or its image to tick it, and tap Done (or press Esc) to leave selection mode without changing anything.
The confirmation tells you exactly what you are about to lose, because this cannot be undone.
Large clean-ups are processed in batches with live progress, so a slow connection never leaves you guessing.
Language and regional preferences for your dashboard.
Your brand and copyright details for generated content.