Model-agnostic production · Private beta

Generative AI production,
connected.

Create images, video, and audio with your choice of supported AI models. Switch engines within reusable canvas workflows, then carry the work through versioned takes, team review, and editing.

And Action! / CanvasRoom to think. Tools to make.
Character reference images connected to a Canon Image Set, beside the Assets browser in And Action!. Open full view
Inside And Action!Actual application screenshots · Private beta

01 / Your models. Your workflow.

Change the model.
Keep the workflow.

A great generation is only the beginning of a production.

Teams still need to compare versions, trace how a shot was made, share feedback, and bring it into the cut. And Action! connects those steps around the work, whatever supported engine you choose.

Choose the model that fits the shot

Same canvas.
Another creative possibility.

Select a different supported engine on your generation node without rebuilding the surrounding graph. Adjust the model's settings, generate another take, and compare the results.

Explore model choice

Selected supported video models

  • ByteDanceSeedance 2.5
  • MiniMaxMiniMax H3

Connect your own provider API keys. Available models, inputs, and controls depend on the connected provider.

  1. 01

    Keep the connections.

    Build generation into a reusable workflow with references and finishing steps.

  2. 02

    Keep the versions.

    Compare saved takes as you explore different models and creative choices.

  3. 03

    Keep the route back.

    Reconstruct the exact saved canvas behind a take version and build on it.

02 / The connected workflow

Keep the creative thread.
From first idea to final cut.

The canvas is your home base. Your shots, versions, feedback, and edit stay part of the same production.

Canvas

Make room
for the idea.

Bring references, storyboards, and shot workflows together on an open canvas. Choose AI engines per generation node, connect the steps, and follow the work as it runs.

  • Visual planning and connected workflows
  • Shot composition and cinematic controls
  • Shared editing and comments in context
Character reference images connected to a Canon Image Set on the And Action! canvas.Open full view ↗

Takes & provenance

Reopen the workflow.
Keep creating.

Choose a specific take version and reconstruct the exact saved canvas workflow that created it. See how that result was made, then use the workflow as the starting point for your next iteration.

  • Restore the workflow behind a specific take version
  • Inspect its nodes, connections, inputs, and settings
  • Continue creating from the restored canvas
The And Action! Takes workspace showing saved production outputs.Open full view ↗

Review

Bring everyone
into the picture.

Look at the work together, or leave feedback for the next time zone. Keep the conversation attached to the shot and the frame.

  • Frame-by-frame playback and annotation
  • Live review sessions and async notes
  • Feedback where the work lives
A production shot open in the And Action! Review workspace.Open full view ↗

Editor

Find the rhythm.
Build the cut.

Take the work into a built-in timeline editor. Shape the sequence with video, audio, effects, and grading, then export the result.

  • Multitrack video and audio editing
  • Transitions, effects, and color tools
  • Export and editorial handoff
The And Action! timeline editor with a production sequence and preview monitor.Open full view ↗

Workflow templates

A starting point
for your next idea.

Start with a ready-made workflow for a film, character, storyboard, or finishing task. Open it on your canvas and make it your own.

  • Browse by what you want to make
  • Connected nodes, ready to customize
  • Free and paid requirements shown up front
See an example workflow ↗
The actual Workflow Templates browser, with categories for video, images, characters, storyboards, editing, VFX, and audio.Open full view ↗

03 / Production provenance

From a take
back to its canvas.

Choose a specific take version and reconstruct the exact saved canvas workflow that created it. The nodes, connections, inputs, and settings come back together, ready to inspect and use.

Which model made this take—and how?

Follow the restored workflow from its inputs to the take. Inspect the prompts, settings, and recorded engine/model details behind that version.

Make it your next starting point.

Work directly with the reconstructed canvas. Adjust a prompt, swap an input, or extend the workflow to explore the next iteration.

Carry the record forward.

Export a lineage bundle with the graph, recorded execution details, parameters, output references, and job records for inspection and handoff.

The canvas restored from take HIO0020 v019, showing character references, voice processing, the prompt, video generation, color grade, and Submit Take. Open full view ↗
Actual workflow reconstructed from HIO0020 · v019. See the take-to-canvas walkthrough ↗

Project teams & access

Your project. Your team. Defined access.

Every project has its own team and gated access. Set who can participate with admin, editor, and viewer roles, and scope access to specific sequences, shots, or assets.

04 / A creative partner

An agent that
works with you.

Describe the outcome. Build it together.

Ask for help, plan a workflow, or have the agent build and connect nodes on your canvas. The work stays visible and editable, so you can step in and direct the details.

It asks for approval before paid operations. You stay in control of the creative decisions and the spend.

From a connected idea to a working result

Choose what runs.
Keep the process in your hands.

Run one node, everything it depends on, or the whole connected workflow. Reuse finished results as you iterate, then turn the work you want to keep into versioned takes.

A selected Submit Take node with the Run menu open, showing upstream, downstream, connected graph, result reuse, and mint takes options. Open full view ↗

05 / More ways to make

Follow the idea.
Choose the tools as you go.

Mix media and generation engines in the same workflow, with creative controls beyond the prompt.

01

Image & video

Generate, extend, modify, reframe, and refine shots with the engine that fits the job.

02

Voice & sound

Bring voice, music, and captions into the production alongside your pictures.

03

3D & shot direction

Explore assets and camera tools. Shape framing, movement, and visual intent.

04

Cleanup & finishing

Work with masks, object removal, upscaling, and grading as connected steps.

Explore the detailed feature tour

Built from production experience

Creative freedom.
Production discipline.

Built on 30 years of feature-film visual effects experience, including work at ILM, Weta Digital, DreamWorks, and DNEG. And Action! brings the exploration of generative tools into a connected production workflow.

Studio conversations

Let's talk about
what comes next.

Interested in the product or the thinking behind it? Get in touch about creative technology roles, studio partnerships, or strategic opportunities.

Currently in private beta. Access is limited while the studio is in active development.

Studio conversations

Let's connect.

A role, a collaboration, or an idea for what comes next. I'd love to hear what you have in mind.

Message delivery is unavailable until this form is connected.