Models & pricing

Capy bills for two things: model tokens and machine server time. Usage draws from one balance, in dollars, shared by your whole organization. A subscription plan includes a set amount of usage each month at a discount. You can also add balance directly (minimum $5) and enable auto-reload.

​

Plans

Three plans, sized by the monthly credit grant. Annual billing is 12 months upfront at 20% off, with the same credits granted monthly. Capy Max comes in $100 steps from $200 to $1,000. Every step grants 110% of its price in credits, so $500/mo buys $550 of credits. Try Capy Lite for $1. Your first 7 days cost $1 with a verified card, then $20/mo. One intro per card and per organization. Monthly Lite only. If you use your credits up mid-month, top up manually or with auto-reload.

​

No seats

Every plan includes unlimited members. There is no per-seat price, no included-member limit, and no charge for inviting anyone. Plans size usage, nothing else.

​

Models

Every model appears once in the model picker, named by the company that makes it: openai/gpt-6-sol, anthropic/claude-opus-5. What a model runs on depends on your connections: the first subscription or key in your list that covers it, or Capy credits if none does. Prices are per 1M tokens and apply to usage on Capy credits. Usage on a subscription or key costs nothing from your balance. A ✓ in the BYOK column means you can run the model on your own key from that provider.

​

Subscriptions

A connected Codex, GitHub Copilot, or SuperGrok subscription covers these models, and usage counts against that plan’s limits instead of your balance. Azure runs through an organization account. Connect one under connections.

​

Fast mode

Models with a Fast toggle run through the provider’s priority tier at a higher price. When no fast host is available, the request serves at the standard tier and bills at standard rates. Fast rates per 1M tokens:

​

Reasoning effort

Most models expose a reasoning effort, from none up to max. Each entry supports its own subset, shown in the picker. An unsupported effort fails before the request is sent. Capy never quietly approximates it with a different setting.

​

Connections

Out of the box every model runs on Capy credits, the balance your organization buys. Connect a subscription or an API key and the models it covers run on that instead, at no cost to your balance. A thread uses the first connection in the list that covers its model: yours first, in your order, then your organization’s, then Capy credits.

​

How it works

• One connection at a time. Each message uses the first connection in the list that covers the model: the thread owner’s connections, top to bottom, then the organization’s, then Capy credits. A colleague writing in your thread runs on your connections, never their own.
• Each connection covers a set of models. A subscription covers what its plan includes. A key covers every model from its company unless you uncheck one. Capy credits has the same checklist, so an admin can keep a company’s models off the organization’s balance.
• What you pay. Usage on a subscription counts against that plan’s limits. Usage on a key is billed by the provider. Only usage on Capy credits and machine time come out of your balance.

​

When a subscription or key runs out

A subscription is used up when it hits its plan’s limit, and works again when the limit resets. A brief rate limit is retried and does not count. A connection can also be turned off, a subscription signed out, and a key rejected. When the first connection in the list cannot run the model for one of these reasons, what the thread does next is one setting on the Models page: an admin sets it for the organization, and can let each member set it for their own threads. A key that has no credit left is different: the message fails with the provider’s error, the setting below does not apply, and the key works again once you top the account up where you manage it. When the thread moves on, it posts one note, Codex is used up until 11:08 PM. Running on OpenAI team key., and goes back to the earlier connection as soon as that works again. When it stops, it shows one card with one question. Wait holds the thread until the reset. Use continues on the next connection, and from then on the thread keeps going on its own, Capy credits included. Only the thread’s owner can answer. The same card shows in Slack with the same two buttons, and the CLI prints it with /wait and /continue. If nothing in the list can run the model, the thread stops and says so.

​

Organization settings

Admins manage the organization’s connections and these switches on the Models page; members can see them but not change them.

• Members can connect their own subscriptions and keys, on by default. Off hides members’ own lists and runs every thread on the organization’s connections.
• Members can choose what happens when a subscription or key runs out, off by default.
• In a model’s ⋯ menu in the Models table: Only OpenAI team key (named after the organization’s key for that model) keeps the model on the organization’s keys, never a member’s own connection or Capy credits, for teams with data-handling requirements; Turn off removes the model from every picker in the organization; Compact at sets how full a thread’s context can get before it summarizes its history.

​

Per-task models

Threads and tasks pick models independently. The agent can run a task on a cheaper model for mechanical work, or a stronger one for a hard subsystem. Tasks follow parent thread models by default.

​

When your balance runs out

A thread on Capy credits stops with an out-of-credits card when the balance runs out. Add balance manually or with auto-reload, then send a message; the thread picks up where it left off. A thread on a subscription or key never spends Capy credits without asking first unless your organization set it to keep going; see when a subscription or key runs out.

​

Machine pricing

Each thread runs on its own VM machine, billed hourly by size: Large is the default. You pay while a machine is awake, plus a two-minute wind-down when it goes to sleep. A sleeping machine costs nothing, but waking it to extend its uptime for ports, desktop, file viewer will incur a charge. See machines.

​

Enterprise

Need more than Capy Max’s $1,000/month in credits, custom pricing, or invoiced billing with net terms? We set up a custom plan. Everything else works the same as on standard plans. Contact us →

​

FAQ