Create unlimited workflows.
Pay for live controlled work.
Map, draft, and simulate freely. When Hubson runs live work across your apps, SOPs, approvals, and receipts, usage counts toward your live run capacity. Not priced by workflow count or message volume — invite the people who need to request, approve, and review work. Bring your own AI keys where supported, or see Hubson-managed model usage separately and transparently.
Mapping and trusting before live usage
- Unlimited workflow drafts
- SOP attachment, app/action review, and sample simulation
- Sample-data simulation & sample receipts
- No credit card
First live controlled workflows
- Everything in Explore
- Live app connections & runs
- Approval policy & Action Receipts
- Bring your own AI keys
10–50 person teams making Hubson a weekly operating layer. Best fit after one workflow proves useful.
- Everything in Launch
- Per-workflow budgets & usage controls
- ROI & workflow health reporting
- Browser & code workflows where needed
Standardizing controlled AI execution across functions
- Everything in Growth
- Multi-function workflow rollout
- Org-level access, credentials & budgets
- Priority activation support
Extra capacity: +250 units $149 · +1,000 units $499 · admin-approved, with warnings at 70 / 90 / 100%
A run unit is one live execution
Only live production runs consume capacity. Bigger work costs more units — and the biggest work is scoped, estimated, and approved before it spends.
You can chat freely. Usage counts when Hubson runs live work.
We don't meter the things that make pricing feel unpredictable — messages, workflow count, or model tokens.
Counts toward capacity
- Live production runs using real org context & live app connections
- Controlled actions that produce Activity or an Action Receipt
- Code / browser execution inside a live run
Never billed separately
- Messages, conversations, and approval comments
- Workflow count, seats, or human waiting time
- Model calls, internal action calls, or app actions inside the same run
Find yourself in these examples
Most workflows are Standard runs. Here's roughly where three teams land.
~120 runs / month
Customer escalations and refunds, almost all Standard runs. Launch covers it with room to spare.
A weekly operating loop
RevOps reports, escalations, refunds, and issue handoffs across the team. Growth is the fit.
Browser & code work
Some runs are Extended or Heavy, so units add up faster. Ops gives the headroom.
Answers before you commit
What is a run?
A run is one live execution of a workflow — from the request, through context gathering, SOP checks, app actions, and any approvals, to the Action Receipt. Simulations are free and never count.
What is a run unit?
A unit is how we measure the size of a live run. A Standard run is 1 unit. Extended runs are 3 units, and Heavy runs start at 10 — scoped and approved before they spend.
What counts as Extended or Heavy?
Extended is a larger bounded run — multi-app, technical, moderate batch, or code/test/PR work. Heavy is batch, recursive, browser/code/data-heavy, or long active execution. Open-ended work is blocked until it's scoped.
Are messages or approvals billed?
No. Messages, conversations, approval comments, and human waiting time are never billed. Only live production runs consume capacity.
Are workflows limited?
No. You can create and simulate unlimited workflows on any plan, including the free Explore plan. You only pay when workflows run live work.
Can I bring my own AI keys?
Yes. Bring your own compute and model keys where supported — model usage is shown transparently in every receipt. Where BYOC isn't available, Hubson-managed model usage is billed separately and transparently.
What happens if I hit my limit?
You get warnings at 70%, 90%, and 100%. You can add capacity (+250 units for $149, +1,000 for $499) with admin approval — nothing runs live over your limit without a decision.
What happens if a run fails?
Platform failures do not count toward capacity. Preflight failures and simulations do not count. Production runs that safely complete controlled steps may still consume capacity, and the receipt shows what happened.
Check if you're ready before you pay anything
Run the readiness audit, map your first workflow, and simulate it for free. Move to live run capacity only when the workflow is ready.