Per seat for access.
Credits for the work.
A subscription covers the people who can use Sempleo. A metered pool of credits covers the work agents actually do on their behalf — runs, embeddings, automations. One bill, two dimensions, both visible in real time.
Starter
For a single team piloting Sempleo on its own data. One workspace, shared context, no SSO.
- Up to 25 users
- 12,500 credits / month · +500 per seat
- 5 integrations
- Community support
- 30-day audit retention
Professional
For organizations deploying Sempleo across multiple teams with SSO, governance, and external MCP connections.
- Up to 200 users
- 50,000 credits / month · +500 per seat
- Unlimited integrations
- SSO · SAML / OIDC
- Governance · review queues
- API access
- 3 external MCP servers
- 90-day audit retention
- Priority email support
Enterprise
For organizations that want bring-your-own model, unlimited MCP, and a named point of contact.
- Unlimited users
- 100,000+ credits · custom
- Unlimited integrations
- BYO LLM · Anthropic or OpenAI
- Unlimited external MCP servers
- 365-day audit retention
- Dedicated support
- Annual contract
USD. 17% off annual, billed up front. 14-day free trial on Starter · no card. Non-profits & universities: 50% off.
All the knobs,
by plan.
The ones we
hear the most.
What is a credit?
A credit is the unit of metered work inside Sempleo. Agent runs, embeddings, and automations consume credits. Every action logs exactly what it cost, so you can see credit burn per agent, per team, per user in real time.
Why both per-seat and credits?
Pure per-seat punishes the thing you actually want — smaller teams shipping more. Pure metered makes the bill impossible to forecast. Sempleo charges for the people who have access, and meters the work agents do on their behalf. You get a predictable base and a usage signal you can steer on.
What happens if we exceed the credit pool?
Overage is billed per 1,000 credits at the rate on your plan ($8 Starter, $6 Professional, $5 Enterprise). Alerts fire at 70%, 90%, and 100%; agents never hard-stop mid-month. Most teams run well inside the pool after the first cycle.
Can we bring our own model?
Yes, on Enterprise. Sempleo supports bring-your-own keys for Anthropic and OpenAI — your inference, your bill, your data terms with the provider. Starter and Professional run on Sempleo-managed inference.
What counts as a user?
Any member who can sign in to the workspace. Deactivated users free up their seat automatically, but retain their audit history. Service accounts used by agents are not billed as users.
Is there a free trial?
Yes — 14 days on Starter, no credit card. You get 5,000 credits and up to 10 users to trial Sempleo on your own data. We seed a demo workspace too if you want to see it running before connecting your sources.
How does annual pricing work?
Annual plans are 17% cheaper than monthly, billed up front. Seat adjustments during the year pro-rate. Downgrades take effect at renewal.
How does Enterprise pricing work?
Anchored to seat count, expected credit volume, and support depth. Includes BYO LLM, unlimited external MCP servers, and 365-day audit retention. We quote per engagement during the founding-customer cohort.
Can we start small and grow?
Yes. Most engagements start on Starter or Professional. Seat count, credit pool, and plan tier all move cleanly; contracts re-base at the point of change.
I went back and forth on this. Pure per-seat penalizes the thing you actually want — smaller teams shipping more. Pure metered makes the bill impossible to forecast. So Sempleo is both: you pay for the people who have access, and for the work agents do on their behalf. Every run reports exactly what it cost, and you can watch the line in real time. Founding customers get this fixed in writing before we change it — and we will change it.
