Pricing
Any model on your own keys, plus a little of ours. Needs a Nib account.
$0with an account
Unlimited use of your own keys, with Nib models included.
$16.67/ month
$200 billed yearly
For people who run several agents all day.
$83.33/ month
$1,000 billed yearly
For people whose agents work in parallel all day, every day.
$166.67/ month
$2,000 billed yearly
Shared agents and guardrails for the whole team.
$25/ month per seat
2-seat minimum
Prices in USD, plus applicable taxes.
Enterprise is open to design partners. Talk to us.
From $60 / seat / month · Annual, 25-seat minimum
Your plan's usage at provider prices
Nib models are counted at the model provider's published price for the tokens you use, cache discounts passed through. Usage beyond your plan is charged at list price × 1.05, which includes a disclosed 5% fee.
Your keys, no meter
Bring your own keys on every paid plan, unlimited. Free includes 2,500 agent actions a month on them.
A percentage in the plan, dollars for money
Your plan shows how much of this week you've used and when it resets. Credits, estimates on your own keys and invoices are always in dollars.
Receipts per task
Every Nib-model request is metered by user and model, and you can see it.
No 5-hour windows
One weekly allowance, with no rolling lockouts. In a new account's first 30 days, Nib models have a daily safety cap: $3 on Free, $25 on Pro, $100 on Max, $200 on Ultra and $25 per seat on Team. Credit you've bought isn't capped. Rate limits are only there to stop abuse.
Usage
Your plan includes model usage every week. The meter shows how much of this week you've used and when it resets.
Each week releases 25% of your monthly allowance, so steady use gets about 108% of a month. Unused weeks don't roll over.
Pro typically covers about 40 small coding tasks a week, like fixing a failing test or adding a small function, with Claude Sonnet 5.5, measured on our eval suite (October 2026). How far it goes depends on the model and the task: heavier models and longer tasks use it faster. Max includes 5× Pro's usage and Ultra 10×. On Team, each seat adds about 30 such tasks a week to the team's shared pool. Free and Enterprise count by the month. The allowance doesn't roll over and has no cash value.
Wait for the reset, switch to a lighter model, use your own key, or carry on with credits if you've switched them on. You're told before anything stops.
Our eval suite is a set of everyday coding tasks the agent carries out end to end: fixing a failing test, adding a small function, renaming something across a few files, and similar. For each task we record what it used. On Claude Sonnet 5.5, three in four tasks use about 1/40 of Pro's weekly allowance or less, so Pro typically covers about 40 of them a week; a Team seat's share of the team's pool, about 30. Your own tasks may be smaller or larger than these. Measured in October 2026; we update the figure when the allowance or the model's rate changes.
Credits never expire while your account is open. The details are in Terms §7.
Compare
| Features | Free | Pro | Max | Ultra | Team | Enterprise |
|---|---|---|---|---|---|---|
| Models and usage | ||||||
| Bring your own keys | 2,500 actions / month | Unlimited | Unlimited | Unlimited | Unlimited | + provider allowlists |
| Nib models included | Monthly allowance (efficient models) + $3 welcome credit | Weekly | Weekly; 5× Pro | Weekly; 10× Pro | Weekly, pooled; about 30 tasks a week per seat | Pooled commit |
| Usage beyond the plan | Prepaid credits | Credits (opt-in), limit $50 | Credits (opt-in), limit $200 | Credits (opt-in), limit $400 | Invoiced at list × 1.05, admin cap | Invoiced |
| Hosted actions (connectors, browser, MCP) | Coming soon | Coming soon | Coming soon | Coming soon | Coming soon | Coming soon |
| Gateway rate limit | 60 / min | 300 / min | 600 / min | 1,000 / min | 1,000 / min | 5,000 / min |
| Agent | ||||||
| Desktop app and CLI | ||||||
| Per-model harness adapters | ||||||
| Subagents, skills, hooks, MCP | ||||||
| Checkpoints, rewind, worktrees | ||||||
| Parallel local agents · race lanes | 2 · 2 | 8 · 4 | Unlimited · 4 | Unlimited · 4 | 8 / seat · 4 | Unlimited · 4 |
| Sync and cloud | ||||||
| Synced devices | 2 | Unlimited | Unlimited | Unlimited | Unlimited | Unlimited |
| Cloud session history | 30 days | Unlimited | Unlimited | Unlimited | Unlimited | Retention policies |
| Cloud runner hours on the default runner (coming soon) | 2 to try, once | 20 / month | 100 / month | 200 / month | 20 / seat, pooled | Pooled; VPC runners on the roadmap |
| Self-hosted runners | ||||||
| Nib Crew | ||||||
| Crew members | 3, manual runs | 10 | Unlimited | Unlimited | 10 / seat, pooled | Unlimited |
| Automated runs (schedules, events) | 200 / day | 1,000 / day | 2,000 / day | 2,000 / day | Unlimited | |
| GitHub App (@nib) | ||||||
| Admin and support | ||||||
| SSO | Google and OIDC: Q1 2027 | + SAML, SCIM: on the roadmap | ||||
| Audit log, view and export | 30 days | 30 days | 30 days | 30 days | 30 days | 1 year |
| Support | Community | Priority | Priority | Priority | Priority | |
The workspace, orchestration, sync and compute: synced history, runner hours (coming soon), crew automation and admin controls. Nib models are included every week. Past your plan, credits (if you switch them on) are charged at list price × 1.05, which includes a 5% fee.
As a share of your plan. The plan card shows the percent of this week and this month you've used, and when the week resets. Each week releases 25% of the month, and unused weeks don't roll over. Pro typically covers about 40 small coding tasks a week with Claude Sonnet 5.5; Max includes 5× Pro's usage and Ultra 10×, and each Team seat adds about 30 such tasks a week to the team's pool. Your own keys never count against it.
You choose: wait for the reset, switch to a lighter model, use your own key, or use credits. Credits are off until you switch them on, are charged at list price × 1.05 (including a 5% fee), and stop at a monthly limit you set ($50 by default on Pro).
Active time a session or crew member spends in a Nib cloud runner, billed per second. Billing stops after 60 seconds without activity, and an idle runner goes to sleep after 5 minutes. A runner that’s waiting for your answer stays awake and is billed until you reply. Local sessions and self-hosted runners don’t count. Cloud runners aren’t open yet, so no runner time is counted today.
During the invite-only beta, paid plans aren't on sale yet. Once they are, you can switch or cancel anytime. Plan changes take effect straight away. Moving up charges the prorated difference at once; moving down credits the unused part of your current plan to your next invoice. Cancelling takes effect at the end of the period you've paid for. If a payment fails, your account falls back to Free, and your own keys keep working within the Free allowance.
Download Nib, sign in, paste a key, and ask for something you’ve been putting off.