Pricing
Predictable pricing, built to scale.
Simple, flat-rate plans with enterprise-grade governance on every tier. Self-hosted on day one.
Prices are final for standard deployments; custom infrastructure is scoped in a 30-minute call.
Unlimited workflows and users on every plan · Billed annually · All prices in CHF, excl. VAT
Starter
For teams piloting their first AI workflows.
Billed annually: CHF 8'800 / year, excl. VAT
Tier limits:
- 1 AI agent
- 1 workspace
- Phone, email, Slack support
Growth
Most teams land here.
Billed annually: CHF 26'400 / year, excl. VAT
Everything in Starter, plus:
- Up to 10 AI agents
- 3 workspaces
- Phone, email, Slack support
- 3-day SLA
Enterprise
For regulated, large-scale deployments.
Let's talk
Custom pricing for your scale and security needs.
Everything in Growth, plus:
- Unlimited AI agents
- Unlimited workspaces
- Phone, email, Slack support
- 1-day SLA
- Dedicated AI solution engineer
Compare plans
Everything in every tier, side by side
A complete, unvarnished view of what you get on each plan: no surprise upgrades later.
| Feature | StarterStart with Starter | GrowthStart with Growth | EnterpriseContact sales |
|---|---|---|---|
| Platform | |||
| Workspace members | Unlimited | Unlimited | Unlimited |
| Active workflows | Unlimited | Unlimited | Unlimited |
| Workflow runs | Unlimited | Unlimited | Unlimited |
| AI agents | 1 | Up to 10 | Unlimited |
| Workspaces | 1 | 3 | Unlimited |
| Visual workflow builder |  | | |
| AI & data | |||
| Built-in vector database | | | |
| Document parsing & OCR | | | |
| HYOK (Hold Your Own Key) for external LLM APIs | | | |
| Security & governance | |||
| Self-hosted deployment | | | |
| Role-based access control (RBAC) | | | |
| SSO (SAML / OIDC) | | | |
| SCIM provisioning | | | |
| Audit logs | 30 days | 1 year | Unlimited + export |
| Observability | |||
| Execution traces & step-level logs | | | |
| OpenTelemetry connector (unlimited export) | | | |
| Webhooks | | | |
| Support | |||
| Onboarding workshop | Self-serve | Guided 2-week kickoff | White-glove rollout |
| Support channel | Phone, email, Slack | Phone, email, Slack | Phone, email, Slack |
| SLA-backed response times | 3-day SLA | 1-day SLA | |
| Named customer success manager | | ||
| Quarterly roadmap reviews | | ||
Platform
- Workspace membersUnlimited
- Active workflowsUnlimited
- Workflow runsUnlimited
- AI agents1
- Workspaces1
- Visual workflow builder
AI & data
- Built-in vector database
- Document parsing & OCR
- HYOK (Hold Your Own Key) for external LLM APIs
Security & governance
- Self-hosted deployment
- Role-based access control (RBAC)
- SSO (SAML / OIDC)
- SCIM provisioning
- Audit logs30 days
Observability
- Execution traces & step-level logs
- OpenTelemetry connector (unlimited export)
- Webhooks
Support
- Onboarding workshopSelf-serve
- Support channelPhone, email, Slack
- SLA-backed response times
- Named customer success manager
- Quarterly roadmap reviews
Platform
- Workspace membersUnlimited
- Active workflowsUnlimited
- Workflow runsUnlimited
- AI agentsUp to 10
- Workspaces3
- Visual workflow builder
AI & data
- Built-in vector database
- Document parsing & OCR
- HYOK (Hold Your Own Key) for external LLM APIs
Security & governance
- Self-hosted deployment
- Role-based access control (RBAC)
- SSO (SAML / OIDC)
- SCIM provisioning
- Audit logs1 year
Observability
- Execution traces & step-level logs
- OpenTelemetry connector (unlimited export)
- Webhooks
Support
- Onboarding workshopGuided 2-week kickoff
- Support channelPhone, email, Slack
- SLA-backed response times3-day SLA
- Named customer success manager
- Quarterly roadmap reviews
Platform
- Workspace membersUnlimited
- Active workflowsUnlimited
- Workflow runsUnlimited
- AI agentsUnlimited
- WorkspacesUnlimited
- Visual workflow builder
AI & data
- Built-in vector database
- Document parsing & OCR
- HYOK (Hold Your Own Key) for external LLM APIs
Security & governance
- Self-hosted deployment
- Role-based access control (RBAC)
- SSO (SAML / OIDC)
- SCIM provisioning
- Audit logsUnlimited + export
Observability
- Execution traces & step-level logs
- OpenTelemetry connector (unlimited export)
- Webhooks
Support
- Onboarding workshopWhite-glove rollout
- Support channelPhone, email, Slack
- SLA-backed response times1-day SLA
- Named customer success manager
- Quarterly roadmap reviews
Optional add-on · On-prem AI hardware
Need dedicated GPUs for local models?
Lunnoa packages already run on your own infrastructure. Only add hardware if you want dedicated GPUs for local LLMs - rent yearly via Lunnoa, or buy outright from our onprem.io partner.
Prototyping & POCs: a handful of users testing local AI on small models.
- 1× Strix Halo APU
- AMD Radeon 8060S
- 96 GB @ 0.3 TB/s
- 126 AI TOPS
- ~0.26 KW max
One team running small-to-mid LLMs (up to Llama 70B) in production.
- 1× Blackwell GPU
- Nvidia RTX 6000 WS
- 96 GB @ 1.8 TB/s
- 4,000 AI TOPS
- ~1 KW max
Multiple teams sharing mid-sized models: a department-wide rollout.
- 4× Blackwell GPU
- Nvidia RTX 6000 S
- 384 GB @ 1.6 TB/s
- 16,000 AI TOPS
- ~3 KW max
Frequently asked questions
Still have questions?
Most teams' questions are answered here. For anything specific to your security review, rollout or procurement process, a human is one message away.
Talk to a solution engineer
Get a tailored walkthrough, security review and proof-of-concept plan in 30 minutes.
Yes. Every component (the workflow engine, vector database, document parsing and LLM routing) runs inside your own VPC or on-prem environment. Your data never leaves your infrastructure, and inference is routed to models you control.
Yes. Through our onprem.io partnership you can rent a dedicated GPU server via Lunnoa as a yearly line item, or buy it outright from onprem.io. This is entirely optional: every other capability on Lunnoa already runs on your own infrastructure for free, including with third-party cloud LLM providers.
HYOK (Hold Your Own Key) means your API keys for external LLM providers (e.g. OpenAI, Anthropic, Google) stay entirely in infrastructure you control. Tokens are billed by those providers under your contracts; Lunnoa routes requests without retaining technical access to your keys. HYOK is included on every plan, so you can mix on-premises models with cloud LLMs under your own governance from day one.
Yes. You can upgrade at any time and the new tier takes effect immediately; the prorated difference is applied to your next invoice. Downgrades take effect at the start of the next billing cycle.
All plans are billed annually upfront. The price on each card is shown per month for comparability, but you pay one invoice per year and can add seats or environments mid-term, billed prorated.
No. Every tier includes unlimited workspace members, unlimited workflows and unlimited runs. Higher tiers unlock more AI agents, more environments, and advanced governance.
We run a 2-week guided proof of concept for qualified teams on Growth and Enterprise. You'll deploy Lunnoa on your own infrastructure, build a working agent for one real use case, and decide after.
Listed prices are in CHF and exclude VAT or local sales tax. We can also invoice in USD, EUR, or GBP where your procurement prefers it, and accommodate procurement processes, POs and custom MSAs on Enterprise.
Yes. Prices are final for standard deployments. Custom infrastructure (for example dedicated GPU hardware or non-standard security reviews) is scoped in a 30-minute call and quoted separately.
Für Teams, die Infrastruktur bereits ernst nehmen.
- CTOs und Platform-Teams, die die Passung zur eigenen Referenzarchitektur prüfen
- Security und Compliance mit Fokus auf Datenresidenz und Zugriffskontrolle
- Operations-Teams, die nachvollziehbare Läufe brauchen, keine Black-Box-Automatisierung
- Builder, die unbegrenzte Nutzung unter einer Pauschallizenz wollen