Autodesk has expanded Flow Studio into what it calls an AI-native creative platform, introducing a 3D Editor and Canvas workflow that lets filmmakers direct scenes in three dimensions before using AI to generate the final visuals. The update directly addresses the common frustration that generative AI video tools produce impressive isolated clips but fail when creators need to control timing, camera movement, and performance across a longer sequence.
"Because filmmaking has always been directed in 3D, even though most generative AI video workflows have tried to control it through 2D inputs," said Nikola Todorovic, co-founder of Wonder Dynamics, an Autodesk company. "In the real world, a director works spatially. You direct an actor where to move, where to look, how quickly to perform an action, and how they relate to the camera and the environment."
Instead of relying entirely on text prompts, the new 3D Editor allows creators to block scenes, animate characters, and adjust camera movement in a proper 3D workspace. Once the scene is set, they move into Canvas, a node-based editor, to generate and refine the final 2D output. The two workflows can be used separately or together. "The goal is to give creators flexibility," Todorovic said. "They can stay in 2D when they want speed and simplicity or move into 3D when a shot requires more precise direction."
From performance capture to reusable animation
In a demo, Todorovic started with footage of someone walking. Flow Studio extracted the performance, converted it into animation, and applied it to a 3D robot model. The camera movement from the original clip was also lifted. The robot's animation and the camera move remained fixed while the environment was swapped from a generated city to a sci-fi landscape to an alleyway. The result is a production workflow where individual elements - performance, camera, character, environment - can be reused and directed independently.
"We get quickly to 80% completion of the shot, but getting control in the last 20%, especially around iteration, is where these 2D-only approaches break," Todorovic said. Autodesk's approach separates direction from rendering: creators make spatial decisions first, then let AI handle the final look.
Old skills, new payoff
Autodesk does not see 3D skills becoming mandatory, but it does see them becoming more valuable. "For people who already understand 3D space, skills such as camera blocking, staging, composition, and animation can now have an even greater creative payoff," Todorovic said. The platform is aimed at independent filmmakers, small teams, and studios who need to produce more complex work without a large production infrastructure. "Generative Video AI can now be directed with the same spatial understanding that live-action directors already use," he added.
Autodesk also plans deeper Maya integration, broader asset import, AI-assisted scene construction, and agent-driven directing workflows. The shift reflects a broader industry recognition that AI's value in filmmaking comes not from generating prettier clips but from giving creators precise control over text-to-video results.
Why this matters for creatives
For filmmakers, animators, and visual storytellers, Autodesk's bet is that AI tools should amplify your existing craft skills rather than bypass them. If you can block a scene, time a performance, and frame a camera, you can now direct AI as a rendering engine instead of wrestling with prompts. That means a single artist or a small team can take on work that previously required a full crew - and keep creative control over the decisions that matter.
Your membership also unlocks:
