> ## Documentation Index > Fetch the complete documentation index at: https://docs.heralded.ai/llms.txt > Use this file to discover all available pages before exploring further. # Plans, trials and limits > What counts against your plan, and how trials and content credits work. Your plan sets how much Heralded measures and writes. Prices and each plan's limits are on [heralded.ai/pricing](https://heralded.ai/pricing). Your workspace's **Usage** settings page shows where you stand. ## Trials A trial lasts seven days and needs no card. It usually starts inside Heralded and opens setup immediately. If we cannot determine your tax location, Stripe Checkout opens first to complete the trial, with no card required. It reads your brand the way a paid plan does, with up to 20 prompts across all your brands and up to three seats. All six standard engines stay on. At your first charge, your plan's prompt limit applies automatically, so you can track more prompts immediately. A Plus trial includes one content credit. The rail shows how many days are left in your trial. During your trial, **Billing** shows your plan, trial end date and the price after the trial in one line. It shows **No card yet** and **Add a card** until you add one. Adding a card lets your subscription bill at the end of the trial. Without one, your trial cancels and your brand locks immediately. We keep your brand, prompts, competitors and history. Choose a plan and add a card to resume daily updates on the same brand, without setting it up again. When your trial starts, the Owner receives an email confirming it is active. With a card, it states when billing starts and at what price. Without a card, it asks you to add one before the trial ends. While founding places remain, it names the offer. When you add a card from **Upgrade now** or claim the founding offer during the trial, the Owner receives an email confirming it, with what happens when the trial ends. Adding a card from **Billing** sends no confirmation. When your first invoice is paid, the Owner receives an email confirming the subscription is live. If that first charge fails, the email arrives once a payment goes through. Three days before your trial ends, every Owner and Billing member receives an email linking to **Billing**. Without a card, it asks you to add one to keep your daily updates and weekly reports. With a card, it states the trial end date and subscription price and reminds you that you can cancel before then. If you have already canceled, it says your trial won't renew and nothing is charged. If your trial ends without a card, or you cancel it, you receive an email confirming that tracking and tasks are paused and your setup is kept. If an agency takes over billing, tracking continues and this trial-ended email stays unsent. A paid cancellation keeps your reports readable for 30 days after the paid period ends. A trial that ends without a card has no read-only window. Starting a trial uses no discount or promotion code, including when you arrive through a founding-offer link. An eligible founding discount saved with your card applies when billing starts. The trial line under the brand menu shows days left and **Upgrade now**. While founding places remain, it shows the offer. After the trial, it shows **Your trial has ended** and **Choose a plan**. On a phone, open the account menu to see your trial status. Both buttons open a centred plan dialog on your current page. Choose Core or Plus and Monthly or Yearly, then **Add a card** during a trial or **Subscribe now** after it ends. During a trial, Stripe securely saves your card. Heralded saves the selected plan and any eligible founding discount on your existing subscription. Your trial continues to its scheduled end, when billing starts. Nothing is charged before then. Leaving the card step without completing it changes nothing. The founding offer appears for monthly billing only. After a trial ends without a card, you land on **My Snapshots**, even if you have not added a brand yet. The notice and locked sections offer **Choose a plan**, which opens the same plan dialog. Locked sections always show a centred card over blurred, inactive page content, including when you open a link directly or reload the page. If you have no brand yet, choosing a locked section opens the card over your current page. Navigation stays usable so you can return to Snapshots or Settings. If you have never had a trial, the card offers **Start free trial**. A paid cancellation's read-only notice and archived workspace page also offer **Choose a plan**. On an archived workspace, the checkout restores that workspace and requires an Owner or Billing member. Subscribing again restores everything you set up. After checkout, you may see **Setting up your organization** while your subscription activates and an archived workspace is restored. Heralded checks for up to a minute, then asks you to wait for the confirmation email if activation is still pending. While the founding offer is available for the first 50 signups, the sidebar, monthly plan chooser and **Billing** show it. On Core or Plus monthly, **Claim founding offer** secures a place when you add a card. Your trial keeps running and nothing is charged until it ends. You can enter the public code in Billing. Yearly plans cannot claim it. You can skip the trial with **Or subscribe now** in the plan chooser or Billing. Checkout requires a card and bills you today. A monthly subscription applies the founding offer automatically while places remain, provided your workspace has never claimed it. Yearly subscriptions carry no founding discount. If you arrive through a founding-offer link for an eligible monthly subscription and the coupon is unavailable, checkout states that you will pay the full price shown. The same notice appears if the coupon becomes unavailable while checkout opens. If your trial has ended, choosing a monthly plan also applies the offer while places remain. Canceling after an early claim does not return the place. Billing shows your discounted price after a claim. See [current offer terms](https://heralded.ai/pricing). **Usage** and **Account settings** show your Snapshot balance with **Buy more**. Snapshot credits are separate from the content credits below. [Snapshots](/concepts/snapshot) explains what each one includes. ## Prompts Your plan includes a number of prompt slots for the whole workspace. A prompt is a question you track in AI answers. Each active prompt in your current prompt set takes one slot per live [segment](/concepts/segments), so the same question tracked in two segments takes two. Paused and archived prompts and prompts in retired segments take none. Historical answers take no slots. A brand whose updates are paused keeps its own count but spends no workspace slots. Resuming that brand needs enough free slots for its prompts. Slots are shared across your brands unless you reserve some for one brand. Free slots exclude every reservation and the active prompts of brands sharing the pool. If you can see brand data, **Usage** and **Billing** show how many prompts you use out of your limit. If your role is Billing, the Billing page shows **Usage unavailable** because you cannot see brand data. **Usage** also shows each brand's count and how many slots are free. With billing access, **Buy more** beside each Usage meter opens a centered dialog. The prompt block adds ten slots to your subscription at its existing billing interval. The dialog shows the price for that interval and states when a purchase bills the remaining period, prorated. If your workspace pays directly, prompt usage stops at its available slots. Buy a prompt block or upgrade to raise the limit. An agency pool can use more than its included amount; that usage is charged on the agency's invoice. Every prompt is asked of every engine, and the [Claude](/concepts/engines#claude) add-on uses no extra slots. ## Brands and segments Each plan includes a number of [brands](/concepts/workspaces-and-brands#brand). An account an agency funds from its pool can hold up to 10 brands, and they share the prompts the agency allocates to it. Segment limits apply per brand, and every brand's Default segment counts as one. Core includes five tasks a week. Plus has no task limit. A task is one thing you say yes to, counted once when it starts. Revisions, applying, sending and the check afterwards belong to that task. A long piece also spends a content credit. Heralded's own work, setup items and declined proposals do not count. On Core, a week runs Monday to Sunday, UTC. A task scheduled for a week counts against that week. A task you dismiss after it starts still counts in the week it started. Once this week's tasks are used, **Start** schedules the task for the Monday of the first week with a task free, and it starts that day. **Write this** and **Rewrite** in [Content](/concepts/content#topics) refuse the order instead. ## Agent answers Your workspace shares a monthly allowance of 500 agent answers on Core and during a trial, or 2,000 on Plus. An answer in a team thread or private chat uses one. Colleagues' posts, the Herald's weekly report, teammate updates and confirm-card clicks use none. **Usage** shows **Answers used** and the reset date. The allowance resets with your billing period. A yearly subscription resets each month, anchored on the subscription's start date. Once the allowance is used, agents stop answering until it resets. The thread shows "The team has used this month's answers; they come back on \." This notice uses no model call. You can still post and discuss work with colleagues. ## History Search Console, Google Analytics and crawler counts are frozen with each weekly report. Your plan sets how far back you can read them. Core reads the last 90 days and Plus the last 13 months, counted from today. A report older than that reads as unmeasured, including when you pin an API read to it with `as_of`, and its counts are `null`, never zero. Every connection works on both plans, and new reports keep recording. The [funnel](/figures#pages-topics-and-audit) on each page states the limit. Its `window.history_days` is your plan's history, and `window.clamped` is `true` when the report it belongs to is older than that. ## Competitors Each brand tracks up to 10 [competitors](/concepts/competitors). The engines still name any brand they like, and the limit is on how many Heralded measures in full. ## Content credits [Content Studio](/concepts/actions#content-studio) spends one content credit per piece, and one more for each further round of rewrites. Plans with credits top your balance back up each month. Unused monthly credits don't carry over, and credits you buy stay until you spend them. You can reserve your plan's monthly credits for a brand. Its reservation is spent over your payer's billing period, which can start mid-month. A refund returns room to the period of the original spend. Your brand can spend only its remaining reservation and what the workspace balance holds. Writing never uses a prompt slot. **Home** and **Usage** show credits spent over your payer's billing period. Reservation edits and managed-account allocations preview the figures that saving will enforce. A preview does not reserve anything; saving checks the current pool again. **Credits left** is what the current brand can spend, capped by its remaining reservation and the workspace balance. **Content credits** in Billing and **Content credit pool** in Usage show the shared balance across brands. When your plan includes no content credits and your balance is empty, Billing says they are not included on your plan. Credits you buy appear as a balance you can spend. **Buy more** beside the content credit pool opens the same dialog with packs of one, five or ten credits. Choosing a pack and confirming opens Stripe Checkout. Billing links to Usage for top-ups. Snapshot packs hold one, three or ten credits. When the balance is empty, writing stops and the work waits for you. Heralded never buys credits on your behalf. ## API and assistants The [API](/developers) and [AI assistants](/assistants) are included on every plan and don't count against its limits. Read a brand's prompt meter through [`GET /v1/brands/{brand_id}/settings`](/api-reference/v1/get-brand-settings) or the [MCP `get_settings` tool](/connect-mcp). `prompts_used` counts its active prompts, excluding retired segments. A paused brand keeps this count while its prompts spend no organization slots. `prompts_reserved` is its current reservation, or `null` when it shares the pool. A brand grant opens this meter without opening organization usage. Organization members can read the same prompt, credit and agent-answer meters through [`GET /v1/organizations/{organization_id}/usage`](/api-reference/v1/get-organization-usage) or the [MCP `get_usage` tool](/connect-mcp). A grant to one brand alone does not open organization usage. In the response, `content_credits.balance` is the pool and each brand's `credits_left` is what that brand can spend. `agent_answers.used` counts the period's answers and `agent_answers.cap` is the allowance. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.