No two companies start from the same place, so we set Kico up with you: your apps, agents and models. The price follows what we build and how much you use it.
We sit with the people who'll use Kico and learn how your company works before anything is built.
The tools you rent today, rebuilt around your data and your way of working, each with its assistant.
Invoices, reports, follow-ups, monitoring: the recurring work, handed to agents with the right approvals.
Short sessions so everyone knows how to ask Kico for the next app or the next agent themselves.
Cloud models with your keys, your own GPUs, or both, and fine-tuning on your own work where it pays off.
Servers, sandboxes, backups, updates and monitoring, in our cloud or on yours.
When something needs a person, you talk to the team that wrote it.
New apps, new agents, new people: the setup grows, and so does what Kico can do for you.
A call or a demo. You tell us what your team does all day and what you rent tools for.
A one-time setup: your environment, your first apps and agents, your models. Its size is what we build together.
A monthly fee that follows your use: the people, the apps and the work your agents do. It grows only when you do.
We run Kico for you. Your team starts within days.
Kico on your servers. Nothing leaves your building.
| Kico Cloud | Kico Enterprise | |
|---|---|---|
| Where Kico runs | Our cloud (US) | Your servers or private cloud |
| Models | Your keys, plans or own GPU | Your own GPUs only, or as you choose |
| Data leaves your network | To the model providers you pick | Never, with local models |
| Apps, agents, assistant | All | All |
| Weekend training on your work | On your Spark, opt-in | On your hardware |
| Who updates Kico | We do, every week | We do, on your schedule |
Because Kico does whatever your team can describe, and no two companies start from the same place. We set Kico up with you, so the price follows what we build together and how much you use it, like a suit made to measure.
On Kico Cloud you can use your own keys or plans (Anthropic, OpenAI, Gemini, OpenRouter and others), so model use is billed by the provider you pick, or costs nothing on your own GPUs. Kico shows every token per project, agent and person.
On Kico Cloud, often within days of the first call. Kico Enterprise depends on your servers and GPUs; we look at what you have and propose the setup.
Yes. Most teams start with one app and a couple of agents, see what Kico does for them, and grow from there.
Not with local models: prompts, files and data stay on your servers. If you choose to, you can still allow cloud models for some work.
Something else? Ask us or write to hello@kico.dev.
