ToonKit's Product Philosophy (2): Empowering Creators to Reach Their Full Potential
Scoping The Future of the Anime Production, For Creators, By Creators, With Creators
Hello from Team ToonKit.
Generative AI is advancing at an incredible pace. Just as "vibe coding" has transformed software development, AI is rapidly lowering the barrier to entry for video creation.
Yet one thing has become increasingly clear: even when creators use the exact same AI model, the gap in the quality of the final results continues to widen.
Our vision has never been "to make it possible for anyone to create animation with AI." Instead, it is this:
To build tools that enable skilled creators to fully realize their creative potential.
This post is about the thinking behind our recent updates, and the direction Team ToonKit is taking as we continue building the product.
Recreating a Professional Animation Pipeline
One of ToonKit's core features—and one that many of you have embraced—is Create Animation from Scenario.
From the beginning, this workflow was designed around a real animation production pipeline, translated into an AI-friendly user experience.
Creators first establish essential assets such as characters and backgrounds, organize scenes based on a beat sheet and production plan, and then build each shot sequentially from that foundation.
The experience we wanted to deliver was simple.
Not using AI to mass-produce disconnected short clips, but helping creators maintain consistency across longer, story-driven productions.
Recently, however, conversations with several creators producing exceptionally high-quality work using ToonKit challenged some of our assumptions.
"ToonKit's 'Create Animation from Scenario' workflow is incredibly useful when we need to produce high-quality videos within a tight deadline."
It wasn't exactly the use case we originally had in mind—which is precisely why it was so valuable.
That single comment reminded us of two important truths we had overlooked.
First, professional creators never compromise on quality. Even when working under intense time pressure, they insist on maintaining a consistent art style, cohesive assets, and carefully constructed scenes.
Second, experienced creators don't rely on a single workflow for every project. Their process changes depending on the type of production, the scene they're creating, the assets they're working with, and even the circumstances of that particular day.
More Time Than We Ever Expected Is Spent on Assets
Professional creators invest an extraordinary amount of time into assets, reference images, and concept art.
Far more than we originally anticipated.
When we analyzed ToonKit usage data, we noticed that many users were importing character assets they had created elsewhere. Because of that, we've invested significant effort into improving ToonKit's built-in character asset creation tools.
Interviews with professional creators helped us understand why.
Creating a character isn't simply a matter of generating a turnaround sheet or a single character reference.
Instead, creators often:
- Generate dozens of concept variations in Midjourney, refining the character through countless iterations until the design matches their vision. When necessary, they introduce additional reference images to fine-tune anatomy, costumes, rendering style, and countless small details.
- Prepare multiple character sheets so that the same character can appear naturally under different lighting conditions and environments without breaking visual consistency.
- Even create scale references comparing characters with other cast members and important props, ensuring consistent proportions across every scene.
In other words, they design assets as comprehensive creative systems rather than isolated images.
What stood out most was this:
Professional creators already have a clear picture of the scene they want to create. Their priority is maximizing control over AI so the final result matches that vision as closely as possible.
That realization strengthened our conviction to build tools that allow creators like these to fully leverage their expertise.
Why We Built Reference to Video and Canvas
These insights directly led to the rapid development of Reference to Video and Canvas.
Rather than focusing on developer-oriented node interfaces centered around prompts and model parameters, we wanted to create an interface designed for creators—one where completed images and generated videos can be compared side by side through visual intuition.
Canvas was built to maximize the potential of Reference to Video while making complex workflows easier to understand. Connected nodes are visually organized so creators can stay focused on the creative process instead of managing the tool itself.
These features are still in their early stages, and we know there is plenty of room for improvement.
As always, we'd love to hear your feedback. We'll continue iterating at the same speed we've moved so far—and hopefully even faster.
To Our Community
Seven months have passed since ToonKit officially launched.
Our goal is to build a platform where creators can freely choose the workflow that best fits each project, adapt effortlessly to different creative situations, and focus entirely on creating—not on fighting with their tools.
To everyone who has been building alongside us, thank you.
We truly appreciate your continued support.
— Team ToonKit