Reference image creation now happens where the work already is.

With Release 2.0.55, you can generate AI reference sheets directly from the Details panel, attach the result to the selected element, and keep it in the team media library with its original prompt.

You get:

• Generate beside Upload: Create a new reference image from a simple prompt without leaving Studio.
• Attached immediately: The generated image lands on the selected element and appears in the team media library.
• Prompt provenance: Generated media is labeled and keeps the prompt that created it for later review.

Less tab-switching, faster visual direction, and reusable references that stay with your team library.

Create more expressive video directly from Studio with Google Omni Flash.

Release 2.0.57 adds Omni Flash for both image-to-video and video-to-video workflows, with synchronized audio support and optional end-frame guidance for image-to-video generations.

You get:

• New video model options: Generate with Omni Flash from images or existing videos.
• End-frame control: Guide image-to-video results with start and end frames for clearer transitions.
• Smarter length handling: Studio warns when a selected source is too long before credits are spent.

More model coverage, fewer surprise failures, and a cleaner path from source media to finished motion.

Place shots exactly where you want them.

In compose mode, you can now drag Add a Shot into the viewport and drop a new camera directly onto the canvas. The camera follows your pointer while you place it, so building a shot starts from the location you mean, not wherever the current view happens to be.

You get:

• Faster shot setup: Drag a new camera straight into position instead of adding first and adjusting after.
• More spatial control: Drop the shot where your scene composition actually needs it.
• Safe cancellation: Press Escape mid-drag to cancel and remove the new camera cleanly.

Clicking Add a Shot still works the same as before. Dragging just gives you the quicker, more intentional path.

Camera framing just got more precise.

Release 2.0.55 adds a custom camera lens workflow, giving you direct control over lens values, field of view, clipping planes, and near/far helper visibility from the camera lens menu.

You get:

• Preset or custom lenses: Start from a known lens preset or switch into a custom setup.
• Fine camera control: Tune lens, FOV, near, and far values for more deliberate shots.
• Clearer setup helpers: Show near and far helpers when you need to inspect the camera volume.

It is a compact control upgrade for anyone composing shots with specific framing needs.

You can now choose your team at login If you're part of multiple teams, you'll be prompted to select which one you're working in when you sign in. From there, everything opens in the right context — projects, tasks, and views are scoped to the team you selected. No more ending up in the wrong place.

Reference Images for Groups & Sets

Attach reference images to groups and sets, so visualizations stay anchored to the right subject and style. Use a reference image to tell the system: make this room, object, or collection look and feel like this. When you save a group as a set, copy it, paste it, or reuse it later, those references stay attached – so the style direction travels with the work. Why it matters:

• Steer generations toward a specific look, material language, or visual mood

• Keep reusable sets consistent across projects

• Add style context at the group level instead of only describing it in the prompt

• Build reference sheets per group and per entity for more structured visual guidance A good example: attach a reference image to a dining room set, then generate with that visual direction baked in instead of trying to re-explain the style every time.

Onboarding is for everyone

Open Studio and there's a guided tour waiting -- built right into the home screen. Worth checking out whether you're just getting started or you've been in it for a while. The prompt drops you into a ready-made scene, walks through the basics, and gives you a clear path forward: take the tour, learn the workflow, and earn 350 credits to use however you see fit.

What's there when you land

• A guided tour prompt on the Studio home screen

• A ready-made scene to learn with, not a blank canvas

• A clear choice: tour now, or skip for later First sessions that feel like a beginning, not a wall. Zero to shareable scene, faster.

Image generation has a cleaner, composition-first prompting path across supported image models. The new structure puts the 3D scene in charge of space, object form in charge of reality, and style prompts in charge of surface treatment.

The result should be less invention from the model, stronger adherence to the scene you built, and more consistent outputs when generating from the same composition.

Autoprompt now uses Gemini Flash 3.5 for scene understanding. The goal is clearer interpretation of your authored scene, camera, and visible details before prompts are assembled.

Better prompt understanding helps the visualizer describe intent more accurately while keeping the 3D composition as the source of truth.

Learn more in the help docs.

Luma Ray video-to-video support adds another high-quality motion path in the visualizer, including resolution choices where supported.

The backend now handles provider outputs more safely, streaming results into Intangible storage and refunding failed transfers so long video jobs are more reliable.

Learn more in the help docs.

The visualizer now pays attention to the timeline frame you are reviewing. Scrubbing updates scene captures, generated media can remember the frame it came from, and thumbnails are more accurate to the authored shot.

That makes animation-aware image and video work easier to reason about, especially when choosing the exact moment to generate from.

Learn more in the help docs.

Enterprise feature management gives teams more control over which capabilities are available in the app.

It is designed for larger organizations that need rollout control, account-level configuration, and safer ways to manage access to advanced workflows.

The timeline playhead time is now editable directly. Type the exact time you want instead of scrubbing by hand when you need precision.

It is a small control that makes timing, review, and keyframe work feel much less approximate.

Learn more in the help docs.

Parented animation lets connected objects move together through time. When a child object is attached to an animated parent, it follows that motion instead of needing separate keyframes.

Use it for practical animation setups: props carried by characters, objects mounted to vehicles, or grouped scene elements that need to travel as one unit.

Learn more in the help docs.

Snapping now gives you clearer control over placement behavior. Choose Surface when you only want clean placement, Surface + Parent when the placed object should follow what it lands on, or Off when snapping gets in the way.

This makes precise scene setup safer, especially when you are arranging props on animated or grouped objects.

Learn more in the help docs.

Parent Tool Mode gives you an explicit way to attach and detach objects. Use it when one object should follow another, like a prop in a character hand or cargo riding on a vehicle.

The relationship is now easier to create intentionally, without relying on accidental placement behavior.

Learn more in the help docs.

Visualizer playback controls make generated motion easier to review. Scrub, play, and inspect outputs without leaving the visualizer context.

This makes it faster to compare takes, catch timing issues, and decide whether a result is ready to keep or needs another pass.

Learn more in the help docs.

AI Composer can now work with spline-based scene elements. That makes it better at creating authored paths and repeated structures such as roads, fences, hedges, lamp posts, and other elements that follow a curve.

It expands what Composer can build from a prompt while keeping the result editable in the scene.

Learn more in the help docs.

The visualizer now carries scene context more consistently into generation. Scene, style, and lighting information are organized more clearly so the model has a stronger read on what you built.

That means better continuity across attempts and fewer moments where the generated output drifts away from the authored scene.

Learn more in the help docs.

Scene to Video turns your authored scene timeline into a video output, giving you a fast way to review motion, blocking, and edit structure before final renders.

Use it to create greybox animatics from the work you already built in Compose, then share or iterate before spending time on polished visual generation.

Learn more in the help docs.

GPT Image 2 is now available in the visualizer, with support for the image workflows and custom style paths that help you keep outputs closer to a target look.

Choose it when you want another high-quality image model option for composed shots, style-driven exploration, and controlled visual iteration.

Learn more in the help docs.

Seedance 2.0 is now available as a video model option. It gives you another path for turning composed shots and source imagery into motion.

Use it from the visualizer when you want to compare motion style, fidelity, and cost against the other video models in your workflow.

Learn more in the help docs.

The Scene Outliner is easier to scan and faster to operate. The panel now brings scene details, object count, search, sort, selection, pinning, and row actions into one focused workspace.

Groups get the same practical controls as objects, including visibility, lock, edit, duplicate, and delete actions, so large scenes are easier to keep organized.

Learn more in the help docs.

Export Video gives you a clearer path from authored shots to deliverable files. Export at shot, scene, or sequence scope and package the result for review or editorial handoff.

This helps move work out of Intangible when the next step is client review, internal approval, or cutting the sequence in another tool.

Learn more in the help docs.

Buying credits is smoother for teams and high-volume workflows. Bulk credit pack purchases let you add more generation capacity in fewer checkout steps.

It is meant for production runs where you know you will need room for exploration, iteration, and final outputs.

Learn more in the help docs.

Publishing now has a dedicated workflow inside the editor. Use the Publish tab to prepare a shareable project view, choose what to include, and package work for review outside the editing canvas.

It gives clients and collaborators a cleaner handoff path: the right shots, selected media, and feedback context without exposing the full production workspace.

Learn more in the help docs.

Team billing is more flexible. Workspaces can support multiple owners and payment methods, making it easier for teams to share responsibility for seats, credits, and billing continuity.

That means fewer workflow interruptions when an owner is unavailable and more control for production teams managing client work together.

Learn more in the help docs.

Aim Camera and Target keeps your camera locked onto a subject as it moves through the scene. Pick the subject, aim the shot, and let the camera maintain focus while the action changes.

It is especially useful for animation, product moves, vehicles, and character shots where the subject needs to stay framed without constant manual camera adjustment.

Learn more in the help docs.

Animation keyframes now blend more naturally. Pose keyframes interpolate between character states, and transform keyframes have a Smooth option for softer motion paths.

Use it when a camera move, prop movement, or character pose needs to feel continuous instead of stepping abruptly from one authored state to the next.

Learn more in the help docs.

Dark Mode is now live in Intangible Studio.

Working late, in low light, or just prefer a darker interface? You can now switch to a new Dark Mode experience across the app for improved visual comfort and focus.

What’s included

• A full dark theme across core UI surfaces
• Improved contrast and readability in key workflows
• Theme consistency for navigation, panels, and interactive components

How to use it

• Open Settings
• Select Appearance
• Choose Dark Mode (or System to follow OS preference)

Storyboard view is cleaner for larger projects. Scene thumbnails can collapse so you can move through shot groups faster and stay focused on the sequence you are editing.

It is a small navigation upgrade that makes bigger boards less noisy and easier to review.

Learn more in the help docs.

Notifications are easier to manage now. Mark items as read or unread from the notification menu so you can keep important updates visible without deleting anything.

The unread indicator and actions stay tucked into the row, giving you a cleaner inbox for project updates, comments, invites, and account activity.

Learn more in the help docs.

Make fast, targeted fixes to generated images without rerendering the whole shot. Open the pencil tool from a visualizer output, describe the change you want, choose an edit model, and generate a new version from the original image.

This is built for the almost-right render: change a color, remove a distracting object, add a small detail, or clean up a surface while keeping the composition you already like.

Learn more in the help docs.

You've typed "cinematic lighting" enough times. Now you just click it.

Choose a style - photograph, digital illustration, storyboard concept - and pair it with a lighting preset. Misty cobblestone dawn. Stormy coastal tempest. Tropical sunset lagoon. The context window updates automatically. The AI reads the brief. Generate.

You get:

• Presets that mean something: Not labels. Actual guidance baked in - text, imagery, intent - so the AI gets it on the first try.
• Style meets light: Pick any combination. Black and white concept against a golden-hour preset. See what happens. It usually works.
• Bring your own: Preloaded options are a strong starting point. Your style drops right in if you've got one.

Everything in your build, listed. Select, lock, hide, pin, duplicate, edit - all without hunting through the viewport.

Full scene control. Every object in your scene, one panel. Mouse over to select. Right there in Build Mode.

Non-destructive by default. Lock what's done. Hide what's in the way. Nothing moves unless you move it.

New model, same workflow. Runway's latest is now available directly inside Intangible - generate straight from your 3D scene, no export, no tab-switching.

You get:

• 3D-to-video, direct: Your scene is the brief. No describing what you've already built.
• Latest Runway output: Gen 4.5 quality, inside the tool you're already in.
• One less handoff: Build in Intangible. Generate in Intangible. Done.

Less hunting. More directing. Everything you need - projects, teams, tutorials, templates - where you'd expect it.

Create, front and center. The invitation to build is the first thing you see. Because that's why you're here.

Everything in its place. Projects, tutorials, and templates in the main view - each one mirrored in the sidebar so you're never more than one click away.

Teams and settings, untangled. Switch teams, access personal settings, manage admin - all from the top bar. Clean hierarchy, no archaeology.

Notifications that actually work. New notification system. Dismissible. No more badge purgatory.

Type a description. Drop a reference image. Get a 3D mesh. It's in your asset panel now.

Three ways in. Single image, multi-view, or plain text. More angles in, more accurate geometry out.

Non-blocking. Generates while you work. Asset lands in your library when it's ready.

Personal and team assets. Works across every team you're on. Sorted by most recently updated.

You know your start. You know your end. Now every video model does too.

You can now set first and last keyframes across all supported models - that's Kling Pro 2.1, 2.6, 3.0 and Veo 3 - and let the model reason what happens between them. You get:

• Full bracket control: Lock both ends of your shot. No more hoping the model lands where you need it.
• Model-native reasoning: Each model interprets the in-between in its own way - pick the one that fits your shot.
• Consistency that holds: Same scene. Same characters. Start to finish.

Your shot, bookended.

Scene Exports have landed on Intangible, letting you take your worlds beyond Studio. You can now export full scenes as GLTF or USD, bringing you:

• True scene portability: Export complete scenes with clean hierarchy, transforms, and layout intact.
• Cameras that carry through: Every camera exports with accurate position and framing — exactly as you composed it.
• Rights-aware by design: Owned assets export fully optimized, while licensed assets are smartly substituted with high-quality stand-ins.

Build in Intangible. Finish anywhere. Your scenes, now ready to move.

Our biggest release in a year is live!

You told us what mattered most:

• More control over AI generation
• Shot-to-shot consistency
• More image & video AI models
• Cleaner, faster interface
• 3D import & export
• AI prompt assist
• Animations

We listened. This release makes creative direction faster and easier. More control over composition. Less time fighting AI. Deadlines feel manageable.

Try it now and let us know what you think — Discord, email, or chat!

Note: This is a major release and we will be refining features, adding tutorials, and fixing bugs over the next week. Thanks for your patience!

Kling 2.5 and 2.6 are now live on Intangible, expanding our Kling integration to support the latest releases — plus introducing Splines.

This update brings you:

• Latest Kling models: Use Kling 2.5 and 2.6 for improved fidelity, motion, and visual coherence.
• Spline-powered world building: Create flexible roads, paths, fences, and layouts using editable splines — reshape curves, extend segments, and fine-tune placement with intuitive handles.

More control. More flexibility. Same Intangible flow.

Flux Pro and Flux Kontext have landed on Intangible to give image quality a serious upgrade. These new AI models from Black Forest Labs give you:

• Unmatched visual quality & speed: Flux Pro delivers fast iteration, hex-accurate color, cinematic lighting with dramatic light rays — top-tier, raw magnetism in every generation.
• Consistent, controllable storytelling: Flux Kontext keeps characters, scenes, and sequences aligned.
• Precision at the speed of thought: Intangible’s camera control and prompt refinements keep you in flow and edit precisely.

Your visuals, your rules — now faster than ever.

Because friends don't let friends use old AI models - Nano Banana Pro and VEO3 now available on Intangible!

Perfect for:

• Consistent visuals, every frame.
• Bring scenes to life with VEO3 dialogue.
• Pinpoint precision for every output.

Announcing AI Composer – generate 3D scenes and custom environments with a simple prompt.

Perfect for:

• Instantly turning ideas into ready-to-edit 3D scenes
• Rapid prototyping with full control over objects and layout
• Describing anything and seeing it come to life
• Iterating fast, refining details, and exporting images or videos easily
• Saving hours on manual drag-and-drop scene setup

Experience creative world-building—powered by a brand new AI architecture from Intangible that unlocks spatial reasoning by large language models.

Announcing Image to Video Style Transfer – cinematic sequences from any image.

Perfect for:

• Creating full video sequences from a single reference image or key frames
• Applying unified cinematic styles across entire shots
• Easily turning animatics or live footage into polished, stylized videos
• Streamlining your look development and style consistency

Announcing Teams – for seamless collaboration across every project.

Perfect for:

• Instant onboarding and sharing across groups
• Effortless template reuse and asset management
• Visibility into progress across multiple projects
• Frictionless teamwork, every step of the way

See what AI thinks before you generate images or video. Review, edit, and refine before asset generation for precise creative control.

Perfect for:

• Complex scene planning and validation
• Understanding system interpretation
• Fine-tuning creative outputs
• Ensuring prompt accuracy before generation

Save any scene element – objects, configured groups, or entire environments – as reusable building blocks. Drop them into future projects instantly with perfect positioning and settings intact.

Perfect for:

• Intelligent variant creation
• Consistent brand asset deployment
• Team collaboration with shared libraries
• Rapid scene assembly from proven components

Import .glb, .gltf, .fbx, .obj, .dae, and .ply files with automatic size, scale, and orientation normalization. Your assets are instantly ready to use and stored in your library for easy sharing.

Perfect for:

• Custom designs and prototypes
• Real-world product visualization
• Combining your assets with our library