One flat price per organization.
Not per seat. Your bill doesn't grow every time someone joins a standup, and every tier ships the full graph, boards, timeline and the MCP server for your agents.
Yearly is twelve months for the price of ten.
Starter
€0
free forever
Never billed. At a limit, creating pauses. Nothing is charged.
€0
free forever
Never billed. At a limit, creating pauses. Nothing is charged.
For a small team — or you and your agents — getting organized.
Start freeTeam
€29
per month, per organization
Never more than €79 a month. 5,000 work items included, then €10 per extra 1,000, up to five blocks.
€290
per year, per organization — twelve months for the price of ten
Never more than €790 a year. 5,000 work items included, then €100 per extra 1,000 for the year, up to five blocks.
For one growing team: 30 members, 20 projects, sub-projects and custom fields.
Start with TeamBusiness
€149
per month, per organization
Never more than €399 a month. 50,000 work items included, then €50 per extra 10,000, up to five blocks.
€1,490
per year, per organization — twelve months for the price of ten
Never more than €3,990 a year. 50,000 work items included, then €500 per extra 10,000 for the year, up to five blocks.
For an organization running many projects: 200 members, unlimited projects, a year of history.
Start with BusinessFree forever on Starter — no credit card, no time limit. Cancel any time. Paid prices exclude VAT.
Enterprise. Unlimited everything, single sign-on, a contracted SLA and on-premise deployment on your own infrastructure — talk to us.
Puppets: assign work to people who don't need an account.
A puppet is a person without an account — a contractor, a client's reviewer, a freelancer. Assign them work, log time on their behalf, keep every contribution credited to a real name. They never occupy a paid seat, and if they join your organization later, their whole history merges into their new account.
Your agents are never a metered add-on.
Connected AI apps use your plan's normal quotas — there is no AI-credit meter, on any plan. For the AI features Pragma runs for you (meeting transcription, semantic search), each plan includes a monthly allowance. Need more? Plug in your own AI model API key — from Anthropic, OpenAI, Google, Mistral and others, or a model on your own network — and use it without limits, on your own terms.
The full graph on every plan. Here is where they differ.
Every plan includes traceability links, critical path, impact analysis, boards, timeline, phases, time tracking and the MCP server for AI agents. The tiers differ in room and in a few features.
| Starter | Team | Business | Enterprise | |
|---|---|---|---|---|
| What you can build | ||||
| Members | 5 | 30 | 200 | Unlimited |
| Puppets — free assignable people | 5 | 30 | 200 | Unlimited |
| Teams | 1 | 5 | Unlimited | Unlimited |
| Projects | 3 | 20 | Unlimited | Unlimited |
| Plans per project | 1 | 3 | Unlimited | Unlimited |
| Work items | 200 | 5,000 | 50,000 | Unlimited |
| Sub-projects | — | ✓ | ✓ | ✓ |
| Custom fields | — | ✓ | ✓ | ✓ |
| Connected AI apps (MCP) | Unlimited | Unlimited | Unlimited | Unlimited |
| Service accounts — for your own integrations | — | — | 3 | Unlimited |
| Two-factor authentication | ✓ | ✓ | ✓ | ✓ |
| Single sign-on (SAML/OIDC) | — | — | — | ✓ |
| What we run for you | ||||
| Storage | 1 GB | 20 GB | 200 GB | Unlimited |
| Activity history retention | 7 days | 90 days | 365 days | Unlimited |
| Meeting transcription | 30 min/mo | 300 min/mo | 1,000 min/mo | Unlimited |
| Semantic search | 50/mo | 1,000/mo | 10,000/mo | Unlimited |
| Where it runs | ||||
| Pragma cloud, EU-hosted | ✓ | ✓ | ✓ | ✓ |
| On-premise, on your own infrastructure | — | — | — | ✓ |
Questions
- Why flat pricing instead of per seat?
- Because a plan works better when everyone is in it. Per-seat pricing quietly pushes you to leave people out; flat pricing means adding your designer, your contractor or your agent costs nothing to decide. We can price this way because there is no sales team to feed and no per-seat machinery to maintain — and because puppets, who cost us nothing, shouldn't cost you anything either. For scale: ten people on Linear Basic are $100 a month, billed yearly, before any contractor (Linear's published pricing, 18 September 2026). Here that team is €29, and the contractors are free.
- Can I pay yearly?
- Yes. Yearly billing is twelve months for the price of ten: Team is €290 a year, Business €1,490. Overage blocks follow the same rule — €100 per extra 1,000 work items on Team, €500 per extra 10,000 on Business, for the year.
- Are my agents metered?
- No. Connected AI apps have no call budget on any plan. What counts is the work your organization holds — the work items your agents create draw on the same quota as the ones you create by hand.
- What counts as a work item?
- Every item in the graph, whatever its kind: tasks, requirements, deliverables, risks, decisions, questions and the rest. Links between items, comments and time entries are not work items.
- What happens if I go over a quota?
- Nothing is deleted, ever, and you can keep reading everything. On Starter, creating new items pauses until you're back under the limit or on a bigger plan. On Team and Business, work items keep flowing: past the included quota they're billed as overage — €10 per extra 1,000 items on Team, €50 per extra 10,000 on Business, per month, added to your next invoice — up to twice the included quota. Every other quota pauses creation until you're back under it.
- Can I cancel and take my data?
- Yes. Cancel any time from the app; access continues to the end of the paid period, then your organization stays readable for 30 days with self-service JSON export before deletion.
- Who processes payments?
- Paddle, as merchant of record — they handle checkout, invoices and VAT. Prices on this page exclude VAT; Paddle adds any applicable tax at checkout based on your billing country.
- Is my data in the EU?
- Yes — application, database, and file storage run in AWS eu-west-1 (Ireland), with EU-region analytics and email. See the privacy policy for the full processor list and the security page for the controls in force.
- Can we run Pragma on our own servers?
- Yes, on Enterprise. Pragma is available on-premise, deployed on infrastructure you control — talk to us.
Try the whole thing free.
Five members, five puppets, all your agents, and the full graph — no card, no clock.
Start free