Billing
Reactor bills through Stripe: a subscription for the plan, plus a monthly AI allowance per seat. Optional usage-based billing covers anything beyond the allowance, and it is off by default. You will never be charged beyond your subscription unless you turn it on.
Plans
| Plan | Price | What you get |
|---|---|---|
| Free trial | $0 for 14 days | Full access. Your card is not charged until the trial ends. |
| Solo | Free forever | One seat, 20 briefs a month, a $5/mo AI allowance, 5 images or videos, all platforms, MCP access. No card. |
| Team | $99/mo | Everything in Solo uncapped, plus 3 seats, no monthly brief limit, uncapped media (within the AI allowance), team invites, roles, approvals. |
| Scale | $199/mo | Everything in Team, plus 5 seats and a 1-hour setup session. |
| Additional seat | $40/mo each | On Team or Scale, for members beyond the included count. Added when someone joins, removed when they leave, prorated both ways. |
Seat counts include the owner: Team covers 3 people in total, Scale covers 5.
Beyond a plan's included seats, both paid tiers add $40/month per person, so growing never forces an upgrade. On price alone Team stays cheaper at every size; Scale is for teams that want five seats included up front, a larger free-fallback cap, and the 1-hour setup session.
What happens when Solo runs out
Nothing breaks. Past the $5 monthly allowance, text generation keeps working on a slower free model until the month rolls over. Images and videos stop at 5 per month because they carry a per-unit cost with no cheaper fallback. Solo always uses automatic model selection, so the allowance is spent on the best model available and then steps down.
Only the workspace owner can add a paid seat. Managers can invite freely inside the included seats, but a seat that creates a charge needs the person whose card is on file. Inviting past the included count returns OWNER_REQUIRED_FOR_PAID_SEAT.
Manage your plan in Settings → Billing: it shows the current plan, trial days remaining, and the renewal date. Manage Billing opens the Stripe portal for payment methods, invoices, and cancellation.
When you cancel, the subscription stays active until the end of the paid period.
Who can manage billing
| Action | Who |
|---|---|
| Start the first checkout for a workspace | Any member |
| Open the Stripe portal (payment method, invoices, cancel) | Owner |
| Delete the workspace | Owner |
| Turn usage-based billing on or off, set spend caps | Manager and up |
| Set per-member spend caps | Manager and up |
| View team-wide usage | Manager and up |
| View own usage, set own default model, add own BYOK keys | Everyone |
See Team and workspace for the full role matrix.
The AI allowance
Every seat gets a monthly allowance for AI usage:
| Plan | Allowance per seat |
|---|---|
| Solo (free) | $5/mo for the single seat |
| Team, Scale and trial | $20/mo per seat, including the owner |
The allowance is denominated in Reactor-rate dollars: the provider's cost for a call times a 1.5 markup. A $20 allowance covers about $13.33 of raw provider spend, which is several hundred typical generations.
Everything AI draws from it:
- Text: content generation, the assistant, AI edits, insights, and MCP-driven generation. Cost depends on the model; a typical social draft on a standard model is a few cents, and economy models cost a fraction of that.
- Images: about $0.06 per generated image.
- Video: about $0.38 per 5 seconds of generated video, so a 10-second clip is about $0.75.
Allowances are per seat, not pooled. One member exhausting theirs does not affect anyone else's.
When it resets
The allowance resets at the start of each billing cycle, anchored to your Stripe renewal date rather than the calendar month. If you subscribed on the 12th, your allowance resets on the 12th.
Monthly generation cap
Separate from the dollar allowance, plans cap how many pieces a workspace can generate per calendar month: 20 on Solo, 25 on trial, unlimited on Team and Scale. One generation run counts once, however many channels it writes for; blank drafts and imported documents don't count. At the cap, generation is blocked with a "Monthly generation limit reached" error until the month rolls over or you upgrade. Unlike the allowance, the cap is per workspace, not per seat, and resets on the calendar month rather than the billing cycle.
When the allowance runs out
Text keeps working, free. Generation drops to an economy model automatically at no charge, and the model picker locks non-economy models. This free fallback has its own per-period cap, counted across the workspace: $2 of economy usage on Solo, $15 on Team, $25 on Scale, $6 on trial. Past it, text generation stops until you enable usage-based billing, upgrade, or the period resets.
Images and video stop. There is no free fallback for media; generating more requires usage-based billing.
Track where you stand in Settings → Usage.
Usage-based billing
Off by default. Turn it on in Settings → Usage (manager role required) to keep using any model past the allowance.
- Overage is billed at the same Reactor rate as allowance usage (provider cost × 1.5) plus $0.25 per million tokens, metered through Stripe and invoiced in arrears with your subscription.
- The spend cap is a hard ceiling on total AI spend per member per period, allowance included. Defaults: $250 on Team and $400 on Scale. Solo has no overage; it stops at its caps. The maximum settable cap is $1,000 per seat, and $2,000 across a workspace.
- Managers can set a different cap per member from the Usage page or the Team page.
- At the cap, overage stops and text falls back to the free economy model.
Safeguards against surprise charges
- Usage-based billing is opt-in per seat, and only a manager can turn it on. Until it is on, exhausting the allowance never creates a charge; you get the free economy fallback instead.
- Spend caps are enforced server-side on every call, with a $1,000 ceiling no one can exceed.
- If a payment fails (
past_due), overage pauses immediately so a failing card never accrues a growing usage bill. The allowance and the free fallback keep working during dunning. - Your remaining allowance is visible in the model picker and on the Usage page at all times.
Bring your own key
If you add your own provider API keys, text generation runs on your keys instead of drawing the allowance. Solo pays no fees on BYOK usage at all; Team and Scale seats pay a small per-token platform fee. See BYOK.