Pricing
Simple pricing,
no surprises
Free to start. Scale when your team needs more. Your own OpenAI key for LLM costs — you pay exactly what you use.
Starter
For individuals and small teams exploring what their product actually does.
Get started- 1 repository
- Up to 5 features scanned
- Knowledge base with AI chat
- Tracking plan
- Flag plan
- Security audit
- Email support
Team
For teams that need the full knowledge base across their product.
Get started- Everything in Starter, plus:
- Unlimited repositories
- Unlimited features scanned
- Linear integration (ticket drafting)
- GitHub repo source (private repos)
- Experiments & A/B test tracking
- E2E testing plan generation
- /kanon:work — pending changes implemented as review-ready PRs, with vertical slice planning
- Priority support
All plans include: knowledge base, AI chat, security audit, tracking plan, flag plan.
LLM costs are separate — you use your own OpenAI API key.
Questions
How does the free plan work?
The Starter plan is free forever — no credit card required. You get one repository with up to 5 features scanned. It includes the full knowledge base, AI chat, tracking plan, flag plan, and security audit. Upgrade when you need more repositories or features.
What counts as a 'seat'?
A seat is a workspace member who can read the knowledge base and use the AI chat. The Claude Code plugin user who runs scans is one seat. Read-only viewers invited to review the KB also count as seats.
What counts as a 'feature scanned'?
A feature is one unit of your product taxonomy — like 'Credit cards', 'User onboarding', or 'Payment processing'. Discovery proposes features from your code structure; you approve the ones you want scanned. Each scanned feature gets a full guide with all sections.
How much does scanning cost in LLM usage?
On Kanon Cloud, LLM usage is included in your plan — there is no separate API bill and you do not bring your own key. For reference, a typical feature scan costs $1-5 of model usage and discovery costs under $0.50; incremental re-scans are cheaper, because unchanged files are hash-skipped and the extraction cache serves repeated content. If you self-host, you supply your own OpenAI key and pay that usage directly.
Can I self-host Kanon?
Yes. Kanon is a standard Next.js application backed by Postgres, and it runs on any Node.js 22 runtime. Self-hosting is available on both plans; you supply your own OpenAI key and pay that usage directly.
Do you offer annual billing?
Yes. Team is $49 per seat per month, or $470 per seat per year — a 20% saving. Pick the interval at checkout, or switch later from Settings → Billing.
What happens to my data if I downgrade?
Your knowledge base stays intact. On the free plan, you can still read everything that was scanned. You just can't scan additional repositories or features beyond the free tier limits. No data is deleted on downgrade.
Something not answered here? Ask us.