Honest credits — 1 credit = $0.001 of real compute, shown before you spend. Every run uses Opus 5, GPT‑5.6, or Sonnet 5 — no cheap models, no downgrades. Or bring your own key and skip credits entirely.
Students
Free trial → BYO key, or Starter
Every EDU template builds and runs in the sandbox on the 1,000-credit trial. Bring your own key (Moonshot/Kimi, Anthropic, OpenAI) and keep going free — Starter ($14.99) just adds managed headroom.
Individuals
Plus
The magic is the schedule — a Morning Brief only briefs you at 7am if it's hosted. Plus ($19) unlocks always-on agents; a daily brief uses ≈1,020 of its 6,000 credits.
Small business
Plus, or Team per seat
A front desk has to be on shift 24/7 — that's hosted, so Plus ($19) minimum. When staff need shared access, Team bills $15/seat and shares the agents.
Use your own Anthropic, OpenAI, or Moonshot key in the Builder — unlimited, free of our fees; the key is passed straight through and never stored.
One-time credit packs that roll over and never expire — same honest $0.001 credit. Spark 2,500 cr $9.99 · Surge 7,500 cr $24.99 · Nova 25,000 cr $79.99
Invite your team, share hosted agents, and bill per seat. Undercuts the competition — and you only pay for the seats you activate.
Typical estimates from the published token envelopes — your meter always shows the real count, and the exact cost appears before every run. Scheduled agents need Plus or above; every row below fits inside Plus comfortably.
| Template | Build it (once) | Run cadence | ≈ credits / mo | Fits in |
|---|---|---|---|---|
| Morning Brief | 179 cr | Every morning | 1,020 cr | Plus — uses ≈17% of allowance |
| Inbox Triage | 179 cr | Twice daily | 2,040 cr | Plus — uses ≈34% of allowance |
| Research Digest | 179 cr | Weekly | 136 cr | Plus — uses ≈2% of allowance |
| Front Desk | 179 cr | On-shift, daily | 1,020 cr | Plus — uses ≈17% of allowance |
Builds vary by brain: Opus 5 ≈179 cr, Kimi K3 ≈108 cr, GPT‑5.6 ≈210 cr per blueprint. Hosted runs use the Sonnet 5 envelope (≈34 cr incl. the published infra fee). BYO key covers building and sandbox chats for $0 in platform fees — hosted always-on runs are managed and metered on Plus and above.
Slide to your cadence — every figure below is the same live math the meter uses.
Estimated monthly runs cost
≈1,020 credits / mo
30 × ≈34 cr per hosted run (Sonnet 5 envelope, infra fee included). Hosted always-on agents start at Plus — the 1,000-credit free trial covers building + sandbox testing.
One-time build cost — pick your brain
≈179 credits once per agent — BYO key builds for $0 in platform fees.
What counts as a run?
One blueprint generation or one sandbox/agent session. Storage of your agents is always free.
What does “bring your own key” mean?
Paste your own Anthropic, OpenAI, or Moonshot (Kimi) API key and pay the model provider directly — we charge nothing on top, ever. The key is used only for your request: passed straight through to the provider, never stored or logged.
Do verdicts change for paying vendors?
Never. Our editorial firewall is structural: rankings and honest verdicts can't be bought.
Can I cancel anytime?
Yes — plans are prepaid for their period and nothing renews automatically today, so there is no lock-in. Your saved agents stay yours (and exportable) on any tier.
Is annual billing cheaper?
Yes — paying annually saves 20% versus monthly (about 2.4 months free). Switch the toggle above to see annual prices.
How do your credits work?
Honestly. 1 credit = $0.001 of real cost — a published, fixed value. Model tokens are metered at provider list price, and hosted always-on runs add a small published infrastructure fee (4 credits per run + 2 per connector action) so nothing hides in the margin. (Most tools' credits are an opaque unit worth anywhere from $0.0009 to $0.005, and you can't tell.) We show the cost before every run, let you set a hard spend cap, and never downgrade you to a cheap model. Need more mid-month? Top-up packs (from $9.99) add credits that never expire. Or bring your own API key and skip credits entirely.
Which AI models do you use?
Only frontier models — Claude Opus 5 for blueprints, GPT-5.6 for interactive sandbox chats, and Claude Sonnet 5 for hosted automation (you can pin any agent up to Opus/GPT-5.6). No Haiku, no mini, no nano by default. The best model for the job, shown to you.
Prices in USD. Managed inference is metered with plan caps; hosted always-on automation is available on Plus and above and meters a small published hosting + tool fee per run (4 credits + 2 per connector action).