EAS builds, app store submission, RevenueCat, white label
Feature
Builder
Team
Agency
EAS build pipeline
App store submission workflow
RevenueCat integration
White label & custom branding
OTA over-the-air updates
Resource Limits
The real differentiators — capacity that scales with your operation
Feature
Builder
Team
Agency
Active products
1
10
40
User licenses
1
5
15
Collaborator licenses
2
10
30
Monthly AI credits
2K
15K
40K
Concurrent AI agents
2
8
25
Sandbox RAM
8 GB
16 GB
32 GB
Cloud Storage
10 GB
250 GB
1 TB
IncludedNot included—See plan card for value
Enterprise
For organizations that need custom governance, security and capacity.
Custom capacity. Dedicated support.
Custom product, license and committed AI capacity for your organization. SAML/OIDC SSO, SCIM, domain claiming, approved model catalog, BYOK/BYOM, private AI gateway, regional processing controls, dedicated infrastructure, onboarding, SLA, PO and procurement support.
Products
Custom volume
User Licenses
Custom mix
AI capacity
Committed, custom
SSO
SAML/OIDC + SCIM
Models
Private catalog + BYOK
Support
Dedicated + SLA
ProductOS Intelligence
Specialized agents. Frontier models. One shared product memory.
ProductOS routes every stage to the right intelligence while one governed AI pool works across research, strategy, PRD, design, coding, QA, deployment and continuous product operations.
Agent Mesh
Specialized agents coordinate across the product lifecycle instead of forcing every job through one generic chat.
Model Router
Automatically balances intelligence, speed, multimodality and cost for each task, with manual selection when you need it.
Product Memory
Every model works from the same decisions, requirements, design system, code and operating history.
Model freedom, governed by ProductOS
No model-version hardcoding. Availability evolves as stronger models become available.
OpenAIAnthropicGooglexAIMiniMaxDeepSeek
Essential
Fast, efficient models for ideation, classification, summaries and routine edits.
Advanced
Stronger reasoning, coding, multimodal design and long-context product work.
Frontier
Leading models for architecture, high-fidelity design, difficult coding and final review.
Named, role-based access
One workspace. Three access levels.
One shared workspace is allowed. One shared login is not. Every ProductOS license ties to one verified human identity. Activity, approvals and usage stay attributable to a person, not a shared credential.
User License
Full access
Create and edit products
Run build agents and consume AI capacity
Edit source code, PRDs and design systems
Deploy to production
Connect GitHub and export
Edit requirements, backlog and design feedback
Use AI for planning and design feedback
Review, comment and approve invited products
Collaborator License
Contributing
Edit requirements, backlog and design feedback
Use AI for planning and design feedback
Review, comment and approve invited products
Review License
Viewing
Review, comment and approve invited products
Capability
User License
Full access
Collaborator License
Contributing
Review License
Viewing
Create and edit products
Run build agents and consume AI capacity
Edit source code, PRDs and design systems
Deploy to production
Connect GitHub and export
Edit requirements, backlog and design feedback
Use AI for planning and design feedback
Review, comment and approve invited products
ProductOS enforces this with verified named identity, MFA or passkey, reasonable device limits and session management. Risk-based checks flag impossible travel, overlapping productive sessions, rapid device churn and repeated session creation. Step-up authentication and session revocation precede manual review. ProductOS does not rely on IP blocking because legitimate offices, VPNs and remote teams share or rotate IPs.
Products and channels
One product, every surface.
A website, mobile app, desktop client and AI agent are channels of one product identity. They share the design system, backend, context and code. All connected channels are included as part of a product, not separate plan selectors or surcharges.
WebCustom domain, production hosting and backendIncluded
MobileiOS + Android release operations as one channelIncluded
DesktopDesktop packaging and release operationsIncluded
AI intelligenceOne measured credit pool across every ProductOS stageIncluded by plan
Upgrade arithmetic
Why Team becomes the natural upgrade.
Upgrades are based on operating needs, not customer identity. Here is how the math works out.
A small team can stay on Builder
Builder workspace$79/mo
2 extra User Licenses$158/mo
4 extra products$236/mo
Builder total$473/mo
A focused three-person operation can expand without being forced into a team plan too early.
Builder configured to Team capacity
Builder workspace$79/mo
4 extra User Licenses$316/mo
2 extra active products$118/mo
Builder total$513/mo
At this operating size, Team is $64 less each month.
Natural upgrade
Team
$449/ workspace / month
5 User LicensesIncluded
3 active productsIncluded
15,000 AI creditsIncluded
Billing policy
Licenses that flex with your team, not against it.
License additionsProrated immediately when you add a User or Collaborator License.
License reductionsTake effect at renewal, so your team keeps access through the paid term.
Released licensesA paid license released mid-term stays available for reassignment for the rest of that term.
Seat reassignment cooldown30 days, with a support exception for genuine employee departures.
Annual commitmentRecommended for teams. Mid-year additions true-up at renewal.
Usage limitsAdmins set pooled AI and build budgets, with transparent, admin-approved top-ups.
Enterprise provisioningSSO and SCIM automate license provisioning and deprovisioning at scale.
Questions
Predictable by design.
Billing
Is there a free trial?
New paid subscriptions start immediately at the founding launch price - there is no free trial. Founding pricing is available to the current customer cohort and stays locked in while the subscription remains continuously active.
Is there a refund policy?
If something genuinely does not work, email us within 14 days of a charge and we will make it right.
When will I be charged?
The exact charge date and amount are shown before you confirm. If you choose annual billing, you are charged once for the full year on the day you subscribe.
How do I cancel?
Self-service cancellation is available any time before renewal. Your workspace and product data are preserved; paid entitlements remain through the billing term, then the workspace becomes restricted until you renew or export. Reminder emails go out before renewal.
General
Why named licenses instead of a shared account?
One shared workspace is allowed. One shared login is not. Every ProductOS license ties to one verified human identity. Activity, approvals and usage stay attributable to a person, not a shared credential. ProductOS does not rely on IP blocking because legitimate offices, VPNs and remote teams share or rotate IPs. Enforcement relies on verified named identity, MFA or passkey, reasonable device limits and session management, plus risk-based checks for impossible travel, overlapping sessions and rapid device churn.
What is the difference between User, Collaborator and Review Licenses?
Review Licenses can view, comment on and approve invited products, nothing else. Collaborator Licenses add editing of planning artifacts, requirements, backlog items and design feedback. User Licenses unlock the full pipeline: running agents, building, deploying, exporting and managing active products.
How do license changes bill?
Adding a User or Collaborator License is prorated for the rest of your current term. Removing a license takes effect at your next renewal, so the team keeps access through what you already paid for. A paid license you release mid-term stays available for reassignment to someone else for the remainder of that term.
What counts as an active product?
A product becomes active in production once it has a live domain, backend or app store release. Draft projects are unlimited and consume AI credits — only the first production publish or activation consumes an active-product slot. One product can have multiple connected channels (web, mobile, desktop, agent) that all share its design system, backend, context and code, at no extra charge. A product's lineage, including history, design system and code, persists across a fork, clone, new repository, new domain or transfer to another workspace.
Do web, mobile and desktop count as separate products?
No. They are channels of one product identity and share its design system, backend, context and code. All connected channels are included as part of a product, not separate plan selectors or channel surcharges.
What are ProductOS AI credits?
AI credits normalize usage across ProductOS agents, text models and multimodal generation. More complex tasks and more capable models consume more credits. Included capacity and transparent top-ups mean no surprise overages.
Which AI models can I use?
Plans unlock Advanced and Frontier capability tiers from providers including OpenAI, Anthropic, Google, xAI, MiniMax and DeepSeek. ProductOS Auto Router balances reasoning quality, speed, multimodality and cost for each task. Manual model selection remains available where the plan and workspace policy allow it.
What happens when AI credits run out?
New AI builds pause until the allowance renews or an administrator approves a top-up. Products already running in production remain available. Enterprise workspaces can route approved tasks through their own model credentials.
Can an agency transfer a product to its client?
Yes. Agencies can hand off ownership, billing and administration without rebuilding the project. The client continues on its own ProductOS plan, and the agency retains access only if the client invites it back.
Start with one product. Scale with clear capacity.
Every plan includes AI capacity, full product lifecycle and code ownership. One workspace price, with no surprise usage charges.