Pro Studio · Node canvas online

Map creative
AI workflows visually.

Bring image, video, text, voice, control, and automation models into one reusable node graph. Change one step without regenerating the whole process.

Model families40+
Workflow nodes120+
BillingGPU minutes
Yacht promo / reusable flow
38%·auto-save 0:12sRun
Upload Photoimage
SOURCE
Campaign yacht image with product framing and background references.
selected imageready
Prompt Refinertext
brand-safecinematicsummer launch
Turn short creative direction into reusable prompt, negative prompt, and shot notes.
refined textready
Image Generator1024x
VARIANT
Generate hero frames from prompt, source image, seed, and control settings.
images · selected● running
Text to Speechvoice
"Every sunset looks better from the water."
selected audioready
Video Generator4.1s
MOTION
Animate selected frame with camera drift, water motion, and timing controls.
videos · selectedready
Media Muxfinal
EXPORT
Combine video and voiceover into a reusable campaign deliverable.
asset.mp4queued
Creators & studios shipping on Mini AI Studio
◐ NORTHWIND▲ AURORA/LAB▦ PARALLEL◆ KINETIC CO.⌘ MONOFRAME◇ SLOWDIVE⊛ HALO/RIG
Studio features

Built for teams that ship with AI.

The canvas is only one part of the product. Mini AI Studio also keeps sessions, assets, collaborators, GPU runtime, and custom workflows organized around production work.

Session library

Save, duplicate, rename, favorite, search, and organize production flows into folders as the canvas history grows.

manage

Team sharing

Invite collaborators as viewers, runners, or editors, with role-based access and shared session notifications.

team

Media gallery

Keep generated images, videos, and audio in one place with filters, previews, favorites, and download actions.

assets

Parallel generation

Run independent branches faster with queued GPU tasks, node-level status, and targeted reruns from any step.

speed

Custom LoRA workflows

Use character and style LoRAs in repeatable flows, with room for custom training and private model workflows.

custom

GPU-minute control

Track GPU runtime, queue position, active billing sessions, and capacity needs before a large production run.

billing
Node library

One canvas for every generation step.

Build flows from focused nodes for inputs, image generation, video generation, audio, utility steps, batching, and final assembly.

Prompt + Refiner

Prompt inputs, AI rewriting, translation, summaries, keywords, and image-aware prompt refinement.

text

Upload Photo

Drop source images, references, product shots, characters, and masks directly into the graph.

input

Upload Video

Use existing clips as source footage for VFX, head swap, frame extraction, trim, and mux flows.

input

Upload Audio

Route voice, music, or raw recordings into transcription, muxing, and downstream generation steps.

input

Image Generator

Create new images and controlled variants from prompt, pose, depth, and reference inputs.

image

Inpaint

Edit selected regions with masks, prompts, and multiple references without rebuilding the whole asset.

image

Image Upscaler

Upscale selected outputs with sharpness, grain, and quality controls for final delivery.

image

Head Swap

Swap heads in still images while preserving pose, lighting, and garment context.

image

Pose + Depth

Extract pose and depth maps from images so other nodes can reuse structure and camera geometry.

image

Dynamic Camera

Render new camera angles from a source image with azimuth, elevation, and distance controls.

image

Video Generator

Generate motion from image, prompt, camera, and timing inputs inside the same workflow.

video

Video VFX

Apply video effects from source clips and prompts while preserving the surrounding workflow.

video

Video Head Swap

Run head swap workflows on video sources with seed, CFG, pose, and step controls.

video

Frame Extract

Pull frames out of video for image branches, references, previews, and downstream edits.

video

Video Trim

Cut clips into reusable ranges before muxing, VFX, or final assembly.

video

Text Overlay

Add text overlays to video outputs as a graph step instead of a separate editing pass.

video

Music Generator

Generate music from style prompts, lyrics, BPM, key, duration, and language settings.

audio

Text to Speech

Create voiceover clips from copy and route them into video or media mux nodes.

audio

Speech to Text

Transcribe audio to text for captions, summaries, prompt reuse, or script refinement.

audio

Media Split

Split combined media streams into separate audio, video, and image outputs.

utility

Media Mux

Combine video and audio branches into final deliverables without leaving the canvas.

utility

Loop

Repeat node runs across batches, variants, or structured lists of inputs.

utility

GPU Control

Gate heavy generation nodes behind explicit GPU readiness and runtime control.

utility

Comment

Document graph decisions, client notes, and production instructions inside the workflow.

utility
Workflow engine

Turn experiments into production flows.

Most AI tools give you an output and hide the process. Mini AI Studio keeps the process visible, reusable, and editable, so good results can become a system your team can run again.

Keep the recipe

Every asset stays connected to the prompts, source files, model choices, seeds, masks, voices, and export settings that made it.

trace

Change only what matters

Swap a model, adjust a prompt, replace a reference image, or test a new voice without rebuilding the whole creative chain.

rerun

Repeat the win

Use a proven graph for the next product, language, placement, or client request so production gets faster with every finished flow.

scale
Usage-based

Pay for runtime, not seats.

Pro is built around GPU time and workflow storage. Public pricing should stay simple until the docs and billing pages are finalized.

Runtime

$0.17
per GPU minute · $10 / hour
  • Metered to actual runtime
  • No monthly subscription required
  • Use the full Pro node canvas

Large Projects

Sales
for bigger runs, teams, and custom capacity
  • Volume runtime planning
  • Workflow onboarding for production teams
  • Custom capacity for high-volume projects
Contact sales →
Common questions

Still on the fence?

What can I build with Mini AI Studio?+

You can build reusable AI workflows for product images, campaign variants, motion assets, voiceover, editing, and media assembly. The canvas keeps each step connected, so a result is a repeatable process instead of a one-off prompt.

How is this different from a prompt-only image or video tool?+

Prompt-only tools hide the production path. Mini AI Studio shows the graph: source media, prompt refinement, generation, upscaling, voice, video, and export nodes. You can inspect the workflow, change a single node, and keep the parts that already worked.

How does GPU-minute pricing work?+

Runtime is billed at $10 per GPU hour, which is about $0.17 per GPU minute. You pay for the generation time your workflow uses, and larger productions can contact sales for capacity planning.

Can teams reuse the same workflow for future projects?+

Yes. The goal of the Pro canvas is to keep prompts, models, media inputs, and settings together as reusable flows. Teams can adapt a proven graph for the next product, market, language, or campaign variant.

Do I need to know every model before starting?+

No. Start with the outcome you need, then add nodes as the workflow gets more specific: upload media, refine the prompt, generate stills, animate frames, add voice, upscale, trim, or mux the final asset.

What happens when a generation fails or needs changes?+

The graph makes the failure easier to isolate. You can change the node that caused the issue, keep successful inputs and outputs, and avoid spending GPU time rerunning unrelated steps.

When should I contact sales?+

Contact sales if you expect high-volume GPU usage, repeated campaign production, a team rollout, or custom capacity needs. We can help plan runtime, workflow onboarding, and production usage before a large project starts.

Ready to build

Open the canvas.
Build the impossible.

Start in Pro, connect the first few nodes, and keep the workflow when the output works.

Start Building in Pro →Read docs