Three plans, one setup fee, no usage meter. Your agents run on your own API keys, so the model bill is yours directly and never marked up.
Every plan starts the same way. We deploy it, build your first agents with you, wire up your connectors, and train the team. You are live in under two weeks, not after a quarter of onboarding.
Most AI tools bundle model usage into the price and mark it up. We do not. You plug in your own OpenAI, Anthropic, or OpenRouter key, you see every dollar of model spend on your own invoice, and you choose the model each agent runs on.
That is also why agents are included in the plan instead of metered. An agent that is not working costs you nothing. One that is working costs exactly what the model charges.
| Plan | Team | Business | Scale |
|---|---|---|---|
| Monthly | $2,000 | $4,500 | $9,000 |
| Setup and launch | $2,500 once | $2,500 once | $2,500 once |
| People | up to 10 | up to 25 | up to 60 |
| Agents included | 5 | 15 | 40 |
| Extra agents | $75 /agent /mo | $60 /agent /mo | $50 /agent /mo |
| AI usage | Your keys | Your keys | Your keys |
| Hosting | Productive cloud | Our cloud or yours | Your servers or dedicated cloud |
| Connectors (1,400 API-key apps) | Catalog, self-serve | Catalog, set up by us | Catalog, plus custom-built |
| Chat, board, inbox, memory, browser use | Included | Included | Included |
| Agent buildout sessions | Launch only | Quarterly | Monthly |
| Key setup and spend monitoring | Self-serve | Included | Included |
| Dedicated engineer | No | No | 8 hrs /mo |
| Custom modules | No | No | Included |
| SSO and audit log | No | No | Included |
| Support | Email, next business day | Same day, direct line | Same day, direct line, named engineer |
No. Every plan starts with setup and launch, because a workspace with no agents built for your work does not show you anything. Founding teams lock their rate for 12 months instead.
A human with a login. Agents never count against seats; they count against the agent number on your tier.
Anything your key reaches: Claude, GPT, DeepSeek, and open models through OpenRouter. You can set a different model per agent.
Yes. Upgrades take effect the next billing month, and you only pay setup once.
Not out of the box. The catalog is API-key based, and Google, Slack, HubSpot, and Salesforce require OAuth apps that are not in it today. On Scale we build those as custom integrations; on Team and Business the agents can still read and write email through API-key mail services, or work alongside your inbox through the board. Ask us about your specific stack before you sign.
On your servers if you self-host. On our cloud, in a dedicated workspace with your own keys and backups. Nothing leaves without your say-so.
Talk to the team. Scale can be sized up with a second seat block and more agents.
Team size, the work that piles up, and where it should live. We reply within one business day with a plan and a number.
Message the team