Prompt Volume, Seats or Credits: How AI-Visibility Tools Price

Two AI-visibility subscriptions can quote a monthly price while selling different units of capacity. Otterly Standard lists $189/month for a panel of 100 tracked search prompts run daily across four core engines. GeoRankers Grow lists $129/month billed quarterly with ten competitors, three seats, weekly scheduled runs and three on-demand runs a month. Those prices alone do not tell you how much of your own monitoring workload either covers.

This is a guide to six pricing structures illustrated by selected public vendor pages, not a census of every pricing model in the market. The examples below were checked on September 5, 2026. Plan names, limits and offers can change; retain the billing period and entitlement when comparing a quote.

Why a monthly number is not enough

A tool may meter configured prompts, generated answers, projects, competitors, seats or credits shared with content generation. These can coexist in one plan. A configured prompt panel is not a monthly execution quota: 100 questions run daily are different from 100 individual answers. “Six engines” also needs a plan, location and answer surface attached before it describes comparable coverage.

For a simple scheduled panel, an estimate of answer executions is: configured prompts × included engines × runs per day × days. That is planning arithmetic, not a universal billing formula. Repeated samples, locations, retries, missing answers and vendor-specific bundling can change what is counted or charged. Peec’s pricing FAQ similarly distinguishes prompts from AI answers using models and frequency.

Six ways to package the work

1. Prompt-volume tiers. Otterly lists Lite at $29/month for 15 prompts, Standard at $189/month for 100 and Premium at $489/month for 400, with daily tracking and four core engines. API and MCP are separate named features listed from Standard, and additional engines are paid add-ons. The panel-size ladder is easy to see, but features are not held completely constant and the prices are not a straight multiple of capacity. Do not shorten “100 tracked prompts, daily” to “100 prompts a month.”

2. Multi-axis tiers. GeoRankers lists Launch at $41/month billed quarterly and Grow at $129/month billed quarterly. Launch includes five competitors and one seat; Grow includes ten competitors and three seats. Both list weekly scheduled runs, with one versus three on-demand runs a month, and headline allowances of 450+ versus 1,100+ AI answers. These are bundled capacity descriptions, not proof that a buyer can never negotiate an add-on. Compare the actual scheduled coverage and limits rather than equating a seed topic, competitor and prompt. The earlier version’s seed-topic example is not repeated as a current allowance because that field was not established in the reviewed current pricing text.

3. Credit-metered, mixed-cost actions. Rank Prompt lists 150 monthly credits on Starter. Its estimator shows $49/month for Starter, while the annual option shows $39.17/month and $470/year. A scan across its four standard engines together costs one credit; premium Gemini or Grok adds one credit each. An AI-written article costs ten credits. The page’s page-speed-audit example is 150 audits for 150 credits, not the five-to-ten-credit audit assertion previously printed here. Confirm the action definition rather than applying one credit rate to every feature.

For a simple illustration using those published rates, ten articles consume 100 credits, leaving 50 of a 150-credit allowance for one-credit standard scans if nothing else consumes credits. That is hypothetical capacity arithmetic, not a measured customer workload. A fixed subscription can still have a predictable dollar bill: the uncertainty is how much mixed activity fits before a top-up or upgrade. Check resets, expiry, hard stops and overage policy separately.

4. Pay-as-you-go credits. Sleepwalker lists €0.02 per credit, a €5 minimum top-up and no subscription or seat charge. The minimum therefore buys 250 credits, not necessarily 250 identical checks: model-specific probes can have different credit costs. Consult the credit documentation for consumption rules. This can suit intermittent work, but it still requires a volume estimate to compare with a subscription. It is not the only structure for which workload matters, and this article does not claim it is the only pay-as-you-go tool in the index.

5. A broad named bundle plus Enterprise. Gauge lists Growth at $599/month with 600 prompts run daily across six platforms, 18 publish-ready articles and ten seats. Its 108,000-answer monthly figure corresponds to 600 × 6 × 30 in a 30-day example. It also offers custom Enterprise scope. This is the named public plan structure observed for this correction, not a promise that a smaller package will never exist. The current page is inconsistent about extra engines: its table places Claude/Grok under Enterprise while its FAQ says bring-your-own-key access can enable them on any plan. Get that entitlement resolved in writing rather than publishing an unconditional Enterprise-only restriction.

6. Self-hosted software plus services. Canonry presents self-hosting alongside Embedded and Managed offers. Its repository license is FSL-1.1-ALv2: source-available with a competing-use restriction, followed by an Apache 2.0 grant for each version on its second anniversary. Calling the current licensed version simply “open source” or “open-core” obscures that restriction. A self-hosted, bring-your-own-key deployment can involve model API charges, hosting, maintenance and staff time even without a software subscription. Embedded or Managed work can add a separate quoted service cost. Check licensing and operational responsibility as carefully as a SaaS allowance; “free” is not a total-cost estimate.

What to normalize before comparing two quotes

• Define the workload. List the questions, brands/competitors, locations and engines you need, then the schedule and number of repeats. Separate configured questions from generated answers and from charged units.
• Define the action mix. Monitoring, article generation, audits and premium-model calls may consume a shared allowance at different rates. Estimate each action before converting credits into capacity.
• Carry the billing commitment. A monthly-equivalent annual or quarterly rate is not a cancel-any-month offer. Include prepaid totals, taxes where applicable, add-ons and the minimum term.
• Check which limits move together. More competitors might require a plan with more seats and prompts than you need. Ask whether add-ons exist instead of assuming the public bundle is the only possible quote.
• Check what happens at the limit. A hard stop changes delivered capacity; an optional top-up changes spending only when purchased; automatic overage changes the billing risk. A credit system alone does not tell you which applies.
• Include your own operating costs. For self-hosting and BYOK, estimate model bills, hosting, security and maintenance. For any tool, include the person who reviews findings and owns publishing decisions.
• Compare like-for-like outputs. A cheap answer count is not a bargain if it omits the engine, citation detail or export you need. Conversely, paying for an unused content bundle does not improve your monitoring coverage.

FAQ

Why do some tools meter usage as well as seats?

Running scheduled queries and generating content are distinct activities from granting a colleague dashboard access. Vendors can package both into the price. The examples above show usage and seat limits together, but their public rate cards do not disclose complete internal cost structures or establish why the whole market chooses a billing model.

Is a credit-metered tool less predictable than a prompt-count tier?

Not necessarily in dollars. A fixed credit subscription with a hard limit can have a fixed bill while delivering different numbers of actions depending on the mix. Paid top-ups or automatic overages create a separate spending question. Prompt-based tiers also need engine, cadence and add-on limits clarified before their capacity is predictable.

Does “free” or “source-available” mean no ongoing cost?

No. Self-hosting and BYOK can move costs to API usage, infrastructure and maintenance. Source availability is also a licensing question, not a price or permission guarantee; Canonry’s present FSL competing-use restriction and future Apache grant need to be considered separately.

Where should I check a current price?

Start with the vendor pricing page linked beside each example and use its plan-specific terms. The full directory and AI-visibility tools comparison provide further context with their own verification dates. This piece explains pricing structures; it is not an offer or an exhaustive live price list.

September 5, 2026 correction note

Corrected configured prompts versus monthly executions, GeoRankers’ billing-period presentation, Rank Prompt’s action costs and credit predictability, Gauge’s engine-gating certainty, and Canonry’s license/cost description. Removed unsupported claims that these are the six exhaustive market models, that one is most common, or that individual packaging features are unique. Selected public sources were reviewed for these examples; no paid product was run and no historical July/August quote or complete index census was re-verified. Original publication and stored verification clocks are preserved. The modification date records this scoped editorial correction only.