Priced for how you deploy.
Start free on the Agent API, put a team into production on usage-based pricing, or run Cogniseam in your own cloud under an enterprise agreement. Every plan keeps your models, your keys, and your audit trail.
Individuals evaluating the platform
- One workspace
- Agent API + MCP access
- Community support
- Bring your own model key
A team putting governed AI into production
- SSO + role-based access
- Connectors & live capture
- Workflows & approval gates
- Email support
Regulated orgs with security review
- SAML/SCIM, private groups
- Private cloud or self-hosted
- Audit export & retention controls
- Dedicated support & SLA
Shaping the product with us, pre-GA
- Direct roadmap influence
- Hands-on onboarding
- Preferred launch terms
- Named engineering contact
Runs where your data must live.
The same product, deployed to fit your security posture. See the Deployment page for architecture, isolation, and compliance detail.
Managed SaaS on our infrastructure. Fastest to start; SSO, backups, and monitoring included.
Deployed into your own Azure, AWS, or GCP tenant. Your network, your keys, your data boundary.
One process, one data directory, zero native dependencies — runs air-gapped in regulated environments.
What's in each plan.
| Capability | Developer | Team | Enterprise |
|---|---|---|---|
| Workspaces | 1 | Multiple | Unlimited |
| Access control (RBAC) | Basic | Full | Full + private groups |
| SSO / SAML / SCIM | — | SSO | SAML + SCIM |
| Deployment | Cloud | Cloud | Cloud · Private · Self-hosted |
| Audit export & retention | — | Standard | Configurable |
| Support | Community | Dedicated + SLA |
Pricing & security FAQ.
How is Cogniseam priced?
Team plans are usage-based and billed monthly. Enterprise is a custom agreement scoped to your deployment model, seat count, and support needs — talk to sales for a quote.
Can we run it in our own cloud?
Yes. Enterprise deploys to a private cloud in your Azure, AWS, or GCP tenant, or fully self-hosted in a regulated environment. See the Deployment page for the details.
Do we bring our own model keys?
Always. Cogniseam is model-neutral — Anthropic, OpenAI, and Gemini behind one interface, switchable at runtime, with your keys. No data trains anyone's model.
What does the security review look like?
The Trust Center documents every control, honestly labelled implemented / partial / planned. Enterprise engagements include audit export, retention configuration, and a named security contact.
See the value, on your numbers.
Adjust the inputs to your organization. Every figure is derived transparently and is illustrative — a starting point for a scoped conversation, not a guarantee.
Illustrative — an estimate from your inputs, not a guarantee. Talk to sales for a scoped model.
Everything a buying committee needs.
We answer the questions before you have to ask. Ask sales for any of these as a document.
- Implementation timeline & onboarding plan
- Security questionnaire (SIG / CAIQ-style)
- Commercial & licensing FAQ
- Support tiers & SLA
- Budget-planning guidance
- Data-processing & residency terms
- Knowledge sources identified
- Security review owner named
- Deployment model chosen (cloud / private / self-hosted)
- First use case & success metric agreed
- Model provider & key strategy decided
- Executive sponsor confirmed
Governed memory, delivered as business value.
Talk to us about your rollout.
Tell us your deployment model and security requirements, and we'll scope a plan — or a design-partner engagement — around them.
Prefer to start on your own? Create a workspace or apply as a design partner.