Scrunch pricing compared, what the Core plan publishes and the four limits that decide Enterprise

Scrunch publishes one price and one of the fuller feature grids in this market. Read on 31 October 2026, Core is 250 dollars a month with a 7 day free trial, and Enterprise is custom. Core buys 125 unique prompts, 5,000 responses a month and four engines. Four specific limits on that plan are what send a buyer to Enterprise, and all four are published, so you can tell today which side of the line you are on. One more thing to note before you start: www.scrunchai.com/pricing redirects to scrunch.com/pricing, so the company now answers on the shorter domain.

What 250 dollars a month publishes

Published row	Core, 250 dollars a month	Enterprise, custom
Engines	4, named as ChatGPT, Perplexity, Google AIO and Copilot	9, adding Claude, Gemini, Meta AI, Google AI Mode and Grok
Unique prompts	125	custom
Responses collected	5,000 a month	custom
Countries, personas, languages	1, 1, 1	unlimited countries, custom personas and languages
Competitors	5	custom
Site audits and site map	5 a month, 25 pages	full site audits, custom site map
Page optimisations	1 a month	custom
Brand workspaces and licences	1 and 5	custom and custom
Data out	no Looker Studio, no query API, no advanced API	Looker Studio, query API, advanced API, MCP, CLI
Single sign on	Google only	SAML and OIDC
Support	email, self serve onboarding	email and Slack, dedicated account team, review calls

At 250 dollars for 5,000 responses the computed cost is 5 cents a recorded answer, which puts Scrunch in the middle of the vendors that publish both halves. On the same morning Evertune Starter computed to about 1.2 cents, Goodie Core to about 1.8 cents on a daily cadence, AthenaHQ Starter to about 8.2 cents, and our own Growth plan to about 18.7 cents. We are the most expensive of those five per recorded answer and the difference is not a rounding error, because our plans include 20 to 60 pages written and published to your site each month rather than metering alone. If metering is all you want, Scrunch meters more cheaply than we do.

The four limits that decide Enterprise

1. Claude and Gemini are not on Core. Core publishes four engines and neither of those two is among them. On our own plans Claude is included on every plan that reads AI apps, at 3 credits against 1, and Gemini is on all five. If either engine matters to your category, Core cannot answer the question at any volume.
2. One country, one language, one persona. A business selling into two markets has already outgrown Core, regardless of prompt count.
3. Five competitors. That is enough for a tight category and short of what a crowded one needs, and competitor benchmarking is usually the reason a team buys this software at all.
4. Nothing leaves the platform. No query API, no advanced API, no Looker Studio, no MCP, no CLI on Core. If your reporting lives in a warehouse or a BI tool, the integration is an Enterprise line rather than a configuration.

Those four are the honest shape of the plan: Core is built for one brand, in one market, watching a focused prompt set on the four engines Scrunch covers first. Within that shape it is a complete product, and it is self serve, which several rivals are not.

The eight questions the page asks and does not answer

The pricing page carries an FAQ with eight questions on it, including how pricing is calculated, what counts as a prompt, what happens if you exceed your plan limits, whether you can add a user, and whether annual pricing is offered. Read on 31 October 2026, the answers are not in the document the server sends. The questions are in the HTML and no answer text is, in the stripped text or in the raw source.

That matters more here than it would on a marketing page, because four of the eight are the questions a buyer needs answered before a trial: the overage behaviour, the per user cost, the definition of a prompt and the annual term. An assistant asked what Scrunch costs a year cannot read an answer that is not in the page, which is the same defect we found and fixed on our own pricing document, where every currency now prints server side. Ask those four by email and keep the reply.

Where Scrunch beats us, said plainly

The criterion is what the vendor publishes on its own live page that a buyer can check, us included. Scrunch wins four lines. Five user licences on the entry plan, where all of our plans publish one login with full access. Nine engines at the top tier against our five apps. SAML and OIDC single sign on on Enterprise, which we do not publish at any tier. And a self serve 7 day trial of a paid plan, where our own free route is a scan rather than a trial of the paid product.

Where we are the better buy is narrow and worth stating exactly. Claude is inside every plan of ours that reads AI apps, where it is an Enterprise engine at Scrunch. Our ladder is published in six currencies, so a buyer outside the United States can size it without conversion. And our plans include the pages that change the answer, which is a different purchase from a dashboard and is the only reason our cost per answer is higher.

Questions buyers ask about Scrunch pricing

Is 250 dollars a month the real entry price?

It is the published Core figure, read 31 October 2026, with a 7 day free trial in front of it. The page does not publish an annual figure, and it asks the annual question in its own FAQ without answering it in the served document, so request that in writing.

What is a prompt on Core?

The plan publishes 125 unique prompts and 5,000 responses collected a month, so a prompt is the question and a response is the reading. 125 prompts on four engines read weekly computes to about 2,000 responses a month, which fits inside the 5,000 with room for re-tests.

Does Core track Claude?

No. Core names ChatGPT, Perplexity, Google AIO and Copilot. Claude, Gemini, Meta AI, Google AI Mode and Grok appear on the nine engine Enterprise tier.

Can an agency run clients on Core?

Core publishes one brand workspace, so each client would be a separate subscription. The page carries a For agencies view and an Agencies use case, so ask for the agency terms by name. Our page on Scrunch alternatives for a small team or an agency works through what that one published line implies.

What should I check in the trial week?

Whether your own buying questions are in the prompt set, whether the four Core engines are the ones your buyers use, and how many of the 5,000 responses a realistic cadence consumes. Those three decide whether the plan or the tier is the thing you need to change.

A grid you can plan against, and two numbers to request

Scrunch publishes more of its plan than most of this market and the figure is self serve, which together make it one of the easier shortlist entries to price. The gaps are specific rather than evasive: the annual term and the overage rule are asked on the page and answered off it. Get those two by email, convert the quote into cost per recorded answer, then put it against the vendors that publish the same unit, ours included, and accept what the arithmetic says about metering. We are the dearer purchase per answer because our plans also write and publish the pages that move the answer, and if that is not what you want, this is a decision you can settle today from two published documents.

How we know this

We read the assistants rather than guessing what they do. In our own audit of 1,260 saved answers from ChatGPT and Gemini across the United States, the United Kingdom and India, 82.0% of every citation landed on a site the business itself owned, and review directories took 0.7%. Counted as one domain per reading, 16,516 of 20,143.

That is the whole asymmetry: the pages you own are the pages an assistant quotes back to a buyer. See the dated readings and the method on our proof page.

Find out whether assistants name your business, on your own domain and your own buying questions.

By Aditya Goel, Founder, AI Knows Us. Reviewed 1 November 2026.