FredrinSimple, flat pricing
Fredrin is invite-only while we scale, so this is what it costs once you are in: nothing, for a private repo, project memory, and parallel agents on the AI subscription you already pay for. Upgrade to Pro only when you outgrow the free plan.
Free
What your invite unlocks: everything you need to ship solo, with a private repo and project memory included. Nothing to pay, now or later.
- 1 private repo, free
- Unlimited public repos
- Project memory, recall every decision
- Unlimited parallel lanes
- Branch, worktree and session per ticket
- Runs on the subscription you already pay for
- $5/mo of cloud compute included
There is no self-serve signup — an invite is the only way in.
Need more repos or a team?
Upgrade when you outgrow the free plan, never before.
Pro
$20 / moUnlimited private repos and projects, plus team collaboration for everyone you ship with.
Team
$99 / moEverything in Pro for your whole team — 5 seats included, then $29/mo per additional seat. Invite members and share the board.
Enterprise
CustomOrg governance — SSO, an audit log, and role-based access — plus dedicated security review, SLAs, and invoicing for teams that need more than your own subscription.
Every plan includes $5/mo of Fredrin Cloud compute — run a ticket on our hardware instead of your own, free tier included. Your own machine and your own subscription stay the default.
Prices in USD. You bring your own AI subscription — it does the work, with no token markup. Cancel anytime.
What a month actually costs
Two lines on the bill, one of which you already pay — and a third that does not exist.
- Fredrin
- The flat number above — $0 on Free, $20 on Pro — charged per workspace, never per repository and never per ticket.
- Your agent subscription
- Whatever you already pay Anthropic, OpenAI, Cursor, or GitHub. Claude Pro is $20 a month and Claude Max is $100 or $200; a plan you already hold carries over unchanged, because the agent CLIs run on your machine signed into your own account.
- Tokens
- No line at all. Model calls go straight from your machine to your provider, so there is nothing for Fredrin to resell. Running on an API key instead is supported, and that bill is metered by the provider rather than by us.
That second line is the one worth thinking about, because it is where the money is. A flat subscription and a metered API bill diverge the moment you keep more than one agent busy, which is the normal state of a board — four agents working through four tickets burn roughly four times the tokens one does, while the subscription price does not move.
How billing works
- Flat plans
- Free is $0 forever. Pro is $20 a month flat, Team is $99 a month for 5 seats (then $29 per additional seat), and Enterprise is invoiced on custom terms. There are no usage tiers — the price on this page is the whole bill.
- Model usage
- Agents run on the AI coding subscription you already pay for — Claude, Codex, Cursor, or Copilot. Fredrin never bills for tokens and adds no markup on model usage.
- Parallel work
- Running agents is not metered on any plan, so ten tickets building at once cost exactly what one does. The ceiling is your own machine rather than your invoice: the desktop app runs up to ten agents at a time by default, adjustable from one to 64 in settings.
- Cadence
- Plans bill monthly, on the day you upgrade. There is no annual option today, and no per-seat charge until a team passes the seats its plan already includes.
- Starting and stopping
- The free plan needs no credit card. Upgrade only when you outgrow it, downgrade or cancel anytime. Fredrin is invite-only while we scale — free to use once you are in.
- Cloud compute
- Every plan — Free included — comes with $5/mo of Fredrin Cloud compute, for tickets you would rather not run on your own machine. It is metered by the second against that allowance and shown as a live balance in the app; agents on your own hardware stay unmetered, and that is still the default.
Working out which subscription to bring? We did the arithmetic on Claude Max for coding.
Before you pay
- Is there a free trial?
- There is no trial and no money-back guarantee, because the free plan is neither time-limited nor a demo. It is the same desktop app with a private repo, project memory, and unlimited parallel lanes, at $0 with no card on file. Upgrade when a limit gets in your way, not when a countdown ends.
- Do I pay per agent, per ticket, or per minute?
- None of those. Fredrin charges a flat monthly price per workspace, and nothing about how hard you run it changes the invoice — the same $20 covers one ticket a week or a board of them building in parallel all day.
- Does everyone on a team need their own AI subscription?
- Yes. Bring-your-own-subscription means each person's agents run on their own machine under their own login, so each teammate brings the Claude, Codex, Cursor, or Copilot plan they already use. Fredrin's seat price covers the app: the board, the worktrees, the shared project memory, and the review surface.
- What happens to my work if I cancel?
- It stays where it already was. Code lives in your own GitHub repositories and project memory is plain markdown committed alongside it, so cancelling changes what the app will do next rather than what you already have. There is no export step because nothing was ever held anywhere else.
- What am I paying Fredrin for, if my own subscription does the work?
- The surface that runs many of those subscriptions at once: a branch, a git worktree, and an agent session per ticket, project memory that carries decisions between them, and the review and merge path that turns the output into pull requests. The model is yours, the code is yours, and the part being licensed is the board around them.
Questions that are not about money — supported agents, platforms, how Workers stay out of each other's way — are on the general FAQ, and what the app does with your code is set out on the security page.
Free while you're shipping solo. Twenty dollars when you're not.