|
Open beta
No waitlist.
No approval.
No catch. We looked.
Kaitoi is open to everyone. Make an account, claim your free credits, go build something questionable.
Building is free. Running anything that doesn't call a paid provider is also free. When one does, it comes out of your credits at exactly what the provider charges. That is the entire pricing model. There is no clever part.
Start building
|
|
|
We have been heads down, not asleep.
You signed up and then heard very little from us. We would love to blame the algorithm. It was on purpose: Kaitoi is built by a very small team, and every week spent talking about the product is a week not spent building it. So we built. Design partners ran real projects through Kaitoi, hit the walls before you could, and described the experience in detail we did not always enjoy reading. What came out of that is not a roadmap. It is the deeply unglamorous work of making a very wide product hold together under real use.
We are still a very small team. The upside: when you report something, it lands on the desk of the person who wrote the bug, usually the same day. The downside: that person is also doing four other things, and we would rather say so than imply we have a department.
Kaitoi is now in open beta, and anyone can sign up.
We could open with a feature tour. Instead, here are the two things worth knowing before anything else. The tour is right behind them, and fair warning, it goes on a bit.
|
|
Plans and pricing
A run costs what the provider charges. Radical, we know.
Pay-as-you-go is the free plan: no subscription, no monthly minimum, no card held hostage for a trial that renews at 3am. Credits only move when a run actually calls a provider, so anything you build that does not call one runs free, at whatever volume amuses you.
When a run does cost, it comes out of your credits at the provider's own price. No fee on top. No funny-money credits worth forty cents on the dollar. No multiplier tucked into a footnote in a lighter shade of grey. Every charge is itemized per run and per provider. We make money on subscriptions, not on your inference, which is why zero markup is not a launch promotion that quietly expires in March.
Subscriptions buy capability, not a better rate. Indie adds credits that roll over, node and template publishing, MCP and the public API. Studio adds team workspaces, workflow-to-API, raw code and local integrations. Or bring your own keys from Indie up and the provider bills you directly, at the same per-run price. Yes, we are genuinely fine with that.
|
|
|
Privacy
Your work is not our training data.
We never train on your work. We never sell it. We never share it beyond what running the job you asked for actually requires. That is the whole position, not the friendly summary of a longer document that says the opposite.
To be precise rather than merely reassuring: your projects and outputs are stored in Kaitoi, and when a node calls a third-party provider your prompt and your assets go to that provider under their terms, which we do not control and will not pretend to. Your own key changes whose account it runs under, not whether it happens. We tell you which provider handled which run, every time. All of this is in the privacy policy, which we wrote in actual words.
|
|
Studio and Core
One platform, two ways in.
Kaitoi is one product with one set of plans. There are two ways to work with it, and no, this is not a quiz where you pick a side.
Kaitoi Studio is the workspace: a browser-based production environment for people making things.
| › |
Node graph when you want full control, or a simplified workflow view when you just want the pipeline to run and stop asking questions |
| › |
Lab, a conversational moodboard where you talk and the work turns up |
| › |
Storyboard, timeline, and media tools for the deeply uncool part that comes after generation |
| › |
One browser for projects, nodes, files, and outputs, plus a real code editor for when the mood takes you |
| › |
Team workspaces with shared projects, one pooled balance, a weekly cap per member, and every run attributed to whoever ran it, which settles arguments quickly |
Kaitoi Core is the engine Studio runs on: the programmable platform underneath, and the part developers build against directly.
| › |
The execution engine, the sandbox, and the node and agent runtime |
| › |
Managed providers, storage, media handling, and per-run cost accounting that actually adds up |
| › |
The public REST API, the MCP server, security, teams, and billing |
| › |
The CLI and Crew, for the work that belongs in a terminal |
Plenty of people will only ever need one of them. Creatives live in Studio and never think about Core. Developers build on Core and never open a canvas. Both camps are quietly certain they got the better half.
The genuinely fun part is using both: design the workflow in Studio, then ship it as your own application on Core.
|
|
|
Lab
A moodboard you can talk to.
Not every project starts with a graph. Some start with a half-formed idea and a blank wall you have been staring at since Tuesday.
Lab is conversation on one side and a free-form canvas on the other. Describe what you are after and images, video, audio, and 3D land on the board as you go. Mention any item or group by name to refine it, combine it, or spin variations. The whole board stays with the project, bad ideas included, which is how you remember why they were bad.
You can also teach it your own moves. Turn any node or template into a skill, and your pipeline becomes something Penny can reach for mid-conversation.
And Penny is not just forwarding your prompt to a model and hoping for the best.
In advertising it works with Ad Muse, our own creative library. Give it a brief and it pulls real campaign references, works out why they landed, and comes back with original concepts instead of a wall of things that look like other things.
Pick the ones worth chasing and it develops them: scripts, copy decks, art direction, or a full campaign with the hero film, the visual system, social, experiential, and print all hanging off one strategic thread. Then you build the actual assets in the same workspace, which beats emailing a deck to yourself at midnight.
|
|
Nodes
A node can be almost anything.
| › |
AI nodes that call a model |
| › |
Deterministic nodes that are just ordinary programs: parsing, math, file handling, business logic, all the things a model does slower, worse, and for money |
| › |
Agent nodes, where the node itself is an agent with its own tools |
| › |
Plugin nodes that extend what other nodes can do |
| › |
Nodes that are entire applications, interface and all, still wired into the rest of the graph |
All of it is Python.
Write a node yourself, or describe it and let Penny write it while you make coffee. Want a custom interface on top? Add your own JS, HTML, and CSS. Nothing is sealed shut, and there is no store gatekeeper to charm. A node that calls nothing but your own logic costs nothing to run, however many times you run it.
|
|
|
Models and LLMs
A new model ships. One click, and it is a node.
More than 1,400 generative models are available, and image, video, audio, and 3D are only the headline categories. When a provider releases something new, you do not file a feature request and wait for us to notice it. You add it yourself, in one click, and it turns up as a working node.
The catalog keeps up with the industry because you never have to wait for us to keep up with it. We are calling that empowerment.
| Image |
Nano Banana · GPT Image · Seedream · Flux · Krea · Magnific · Recraft |
| Video |
Veo · Kling · Seedance · Happy Horse · Runway · LTX · Grok Imagine |
| Audio |
ElevenLabs · MiniMax |
| 3D |
Tripo · Meshy · Hunyuan3D · World Labs |
| |
and many more, across more types than these |
|
Always the latest version of each. And that is a sample. There are more models than that, more kinds of model than that, and then whatever shipped while you were reading this paragraph.
A number that size is only useful if you can find the right one, so the list is curated and searchable: models, nodes, templates, files, and outputs all in one place, sorted by what they actually do rather than by who shouted loudest on launch day.
Then the language models. Every frontier one, and about 400 more.
Those 1,400 are the models that make things. Language models are counted separately, and there are around 400 of them: the ones that write, reason, and run Penny.
OpenAI, Anthropic, and Google Gemini are supported directly. On top of that, full OpenRouter support puts every other frontier model within reach, along with hundreds of cheaper ones that are quietly excellent and never get a keynote.
| LLM |
Claude · GPT · Gemini · Grok · Qwen · DeepSeek · GLM |
| |
and around 400 in total |
|
Use them in your graphs, or pick which one runs the agent. Expensive when the task deserves it, cheap when it is renaming files.
|
|
Kaitoi CLI and local models
From an idea to a working app, hands free.
The CLI brings Penny to your terminal, where it works on real repositories instead of a canvas.
Next to it sits Crew, which is a different animal entirely: a hands-free, multi-agent delivery engine. Each agent gets a role, planning, implementation, independent testing, review, and none of them is allowed to declare victory, because agents are pathological optimists. The engine verifies the work with real commands and decides for itself when the run is genuinely finished.
Describe what you want, walk away, come back to a working application. Kaitoi-powered, or software that has nothing to do with Kaitoi at all.
Point the CLI at a repo. Get a node.
Give the Kaitoi CLI a GitHub repository or a Hugging Face model and it works out the recipe for you: install steps, environment, run interface, all the parts of the README that were never quite true. The model then runs locally, on your hardware, on your electricity bill.
From there, Kaitoi turns that recipe into a node in your workspace, inputs and outputs already wired, so a local model sits in the same graph as everything else.
That split is the point of putting this in the CLI: the beefy machine stays at home or in the studio doing the work, while you drive Kaitoi from a laptop or a phone, wherever you happen to be.
Already running Ollama, ComfyUI, or Blender? The CLI connects those to your graphs too. And it is not limited to published models: point the generator at your own repository and your own local software becomes a node like any other.
Fair warning: capability varies a lot by model, and the small ones tend to struggle with agentic work.
|
|
|
MCP
MCP in both directions.
Add any MCP server to your workspace and Penny can use it, the same way it uses everything else.
More interesting: Kaitoi is an MCP server itself.
Connect Claude Code, Codex, or any other MCP client to your Kaitoi workspace and let it build nodes, run graphs, and pull results back into whatever you are working in. Your coding agent, quietly operating a media platform behind your back. Available on Indie and Studio.
|
|
Public API
Build your product. We run the machinery.
The public REST API is how developers ship apps powered by Kaitoi. Execution, queuing, storage, provider plumbing, and per-run cost accounting stay firmly our problem.
A workflow you build in Studio can be published as its own endpoint: typed inputs, a URL, and a run your product can call. The graph stops being a canvas and becomes a service.
You build the thing your users actually see, and take all of the credit for it. Genuinely fine by us.
That includes building Kaitoi plugins for the software people already live in.
A panel inside a video editor, a compositing tool, a 3D package, a photo or design app: it sends work to your graphs, Kaitoi runs it, and the result comes back into the timeline, the node tree, or the layer stack the artist is already staring at. No export, no upload, no second window, no third window.
We are not just leaving that to you, either. We are building some of these ourselves, and DaVinci Resolve is first in the queue. It is in progress, not shipped, and we will keep saying so until the day it is not.
The API comes with Indie and Studio. Publishing workflows as endpoints is a Studio feature.
|
|
|
Credits
Free to start, and you can earn more.
Your account starts with free credits, ready to claim. After that, the fastest way to get more is to make Kaitoi better:
| › |
Report a bug we can reproduce |
| › |
Send feedback that changes what we build |
| › |
Publish a community node other people actually use |
We credit your account for all three. Yes, we will pay you to complain about us, provided you complain accurately.
|
|
Still with us? There's more
Project templates · shareable web apps · real-time sessions · connectors and plugins · community nodes and templates · team workspaces and shared budgets · per-run cost tracking · works on mobile and touch.
|
|
|
The closest thing to a catch
This is a beta, and we mean it in the old-fashioned way, not the way that means "shipped, but we would like an excuse." Kaitoi is an ambitious platform: flexibility and adaptability sit right at the core, and a surface that wide takes time to sand down. Closed beta got it solid enough to open. Open beta is where the rest gets fixed, by more people finding the edges we are far too close to see.
So when something breaks, tell us. That is literally what the credits are for.
|
|
Claim your free credits. Go break something.
Open Kaitoi, ask Penny for the thing you actually need, and see how far it gets in ten minutes. Worst case, you end up with a very funny screenshot and some credits for reporting it.
Open Kaitoi
|
|
|
Bugs, ideas, and half-finished work all live on Discord. Come be loud.
|
Questions? Reach us at support@kaitoi.io. A human reads it. Usually the one who broke it.
|
|