Superlines Review 2026: The Agency Plan Beats the Brand Plan

Superlines is a Finnish AI visibility platform that publishes its prices, names nineteen customers, and lists ten engines including Mistral. It is built by Grew Oy in Helsinki, founded in 2020, which makes it older than the category it sells into.

It also has the strangest pricing table in this comparison, and working through it produces a recommendation the vendor almost certainly did not intend.

Disclosure: EchoWi competes with Superlines. Prices, engine lists, customer names and company details below were read from its own pages on 6 August 2026.

The short version

1. From €79 a month, published, with a seven-day trial on every tier. It was ten euros dearer until August 2026.
2. Ten engines exist and you pick three, on the entry plan and on the plan 4.8 times its price alike.
3. Money never buys a fourth engine here. €79, €199 and €379 all say “pick any 3”, and the agency ladder runs at the same three prices.
4. Nineteen named customers, heavily Nordic, which is real proof and also tells you where it fits.
5. No sampling disclosure, like everyone here except Evertune and Siftly.

What it costs, verified

And a parallel agency ladder at the same three prices, which swaps brands for client brands:

Re-read from Superlines’ own pricing page on 8 August 2026. Every tier carries a seven-day free trial, and yearly billing prints €69, €175 and €333 for the same three.

Correction, 8 August 2026. The first version of this table said Agency Growth cost €299 and covered five engines, and that Enterprise covered up to ten. None of that is on the page. The agency ladder runs at €79, €199 and €379, the same as the brand ladder; the comparison row reads “Answer Engines: 3 (choose from all)” on every tier; and the five is client brands, not engines. We turned a brand count into an engine count and built a headline on it. The headline is gone and the arithmetic below is redone.

Publishing prices at all puts it ahead of most of this market before you look at anything else. Bluefish, Brandlight, Peec AI, Evertune and Quattr all route you to a demo instead.

The arithmetic nobody runs

Two ways to divide this table, and they disagree about which plan to buy.

Per prompt, the price falls as you scale, which is the ordinary shape:

But the engine count never moves, so the fairer unit, one prompt tracked on one engine, tells the same story and nothing more:

So the real finding is not a pricing inversion, which is what we published first and which was our arithmetic error rather than the vendor’s pricing. It is this: Growth costs 4.8 times Starter and covers exactly the same three answer engines. Every euro above €79 buys prompts and brands. None of it buys a fourth engine.

That is worth arriving at the call with, because it inverts the usual question. The thing to ask is not what the higher tier adds, it is what to do when three engines are not enough, since no amount of money on this page fixes that.

Ten engines, and you get three

The engine list is genuinely broad: ChatGPT, Gemini, Perplexity, Claude, Google AI Overviews, Google AI Mode, DeepSeek, Copilot, Grok and Mistral.

Two names on it are unusual. Mistral appears on only one other tool we have reviewed, Rankscale, and DeepSeek on two. For a European buyer, or anyone selling to organisations that chose an open-weight or sovereign model deliberately, that combination is not a curiosity.

But the list is a menu, not an inventory. You pick three on Starter and you pick three on Growth, so going from €79 to €379 buys prompt volume and no additional coverage at all. If your reason for upgrading is to see more engines, this is not the upgrade.

That is the same shape as Otterly, where Claude, Gemini and AI Mode are paid add-ons on top of a €29 headline, and the opposite of Rankscale, which puts the same ten engines on its $20 plan as on its $780 one and sells volume instead of access.

The customer list, and what its shape tells you

Nineteen named customers is more than almost anyone here publishes: Tallink Silja, Tietoevry, Finnlines, Elo, Sympa, Sellforte, Trustmary, Dagmar, Elomatic and others, plus Supabase.

Most of them are Finnish or Nordic, and that is worth reading in both directions.

In its favour: naming nineteen real companies is a much stronger signal than a wall of unattributed logos, and shipping companies, pension insurers and IT services firms are not early-adopter names. Somebody is renewing.

Against, or at least worth asking: a customer base concentrated in one small market tells you where the product has been proven and where it has not. If you sell into Spain, Brazil or the Gulf, ask which of those nineteen measure markets like yours, and how the tool handles country and language per run. Our own three-market study found stable citation sets in the United States, Spain and France that did not overlap at all, so a tool proven in one market is not thereby proven in yours.

Supabase is the interesting outlier: a developer-tools company with a global audience, which suggests the product travels further than its customer list alone would imply.

The API and MCP access, which is a real differentiator

Superlines lists an API and MCP access, and it is one of very few tools in this comparison to name the latter.

Why that matters more than it sounds. MCP is the protocol assistants use to call external tools, so exposing your visibility data through it means your own AI assistant can query it directly: ask Claude or ChatGPT how your brand did this week and have it pull the real numbers rather than guess. For a team already working inside an assistant, that removes the dashboard from the loop.

We are biased here and will say so: we ship an MCP interface too, and we think it is the right direction. What we cannot tell you is how good theirs is, because using it requires an account.

Where the reservation is

Not the price, which is published and reasonable. The measurement.

Superlines says it “collects data from the real AI Search interfaces” and “analyzes the real AI Search interfaces your customers see”. Reading the actual interface rather than the API is the harder and better choice, and worth crediting.

What is not stated is how many times. No run count per prompt, no sampling policy, no statement of how a mention is attributed to your brand rather than to a company with a similar name.

That matters because answers vary between identical requests. Asking Google the same question three times in Spain returned eleven cited domains, of which four appeared exactly once. A number produced from one run and a number produced from a hundred look identical in a dashboard.

This is not a Superlines failing so much as a category one: Evertune and Siftly are the only two that publish theirs, at up to 100 runs per prompt per model, and we do not publish ours by default either. Ask it in the trial, where you can watch the counter move.

One number on the homepage we would not repeat. It cites Gartner forecasting that half of traditional search traffic will be replaced by generative AI by 2028. That is an analyst projection about a market, not a measurement, and it is the category of figure we decline to pass on anywhere, including when it favours us.

Where Superlines is the right call

• You are an agency. There is a parallel agency ladder at the same three prices, starting at €79 for five client brands, with white-label reporting, multi-client management and volume pricing above it.
• You want a published price. Rare here, and it lets you budget before a sales call.
• You need Mistral or DeepSeek. Two names almost nobody else covers together.
• You work inside an assistant. API and MCP access mean the data can come to you.
• You sell in the Nordics. The customer list is direct evidence for that market.

Where it is the wrong call

• You want engine breadth cheaply. Three engines at €79 and three at €379. Rankscale lists ten on its €20 plan.
• You are upgrading for coverage. Growth adds prompts, not engines.
• You need to defend a number internally. No sampling disclosure means you cannot say how it was produced.
• Your market is China. No Doubao or Qwen, the gap we built for.
• You need proof in your own market. Ask which customers measure markets like yours before assuming the answer.

What we could not verify

• Run counts and sampling. Not published anywhere we could find.
• Attribution. How the tool decides a mention refers to your brand is not described, which matters most if your brand name is a common word.
• What separates Growth from Agency Growth beyond prompts, engines and price.
• Agency partner pricing. The page promises volume pricing and names no figures.
• The quality of the API and MCP interfaces, which need an account to assess.
• Funding. The site says it is backed by investors and a venture fund and names neither amounts nor firms.

Superlines versus the alternatives

The full comparison covers every tool we could verify, with prices read from each vendor’s own page.

A note on why this review exists at all. Disaggregated against clickstream data, “superlines” is 53 searches a month, the smallest demand behind any review we have written, and 99.5% of its apparent volume belongs to a bus company and a clothing wholesaler. We wrote it because a comparison that skips the vendors nobody searches for is not a comparison, and because the pricing arithmetic above transfers to every tool with a tiered table.

Common Questions About Superlines

How much does Superlines cost?

From €79 a month for 50 prompts across three chosen engines, with a seven-day free trial. Growth is €379 for 150 prompts and still three engines, and a parallel agency ladder runs at the same three prices for 100 prompts and five engines, and Enterprise is custom with up to ten. Prices read from its own pricing page on 6 August 2026.

Which plan is best value?

On cost per prompt tracked per engine, Starter is €0.53, Pro is €0.44 and Growth is €0.42, so the ladder gets cheaper per unit as you climb, which is the ordinary shape. What does not change is the engine count: three on every tier. The useful question is therefore not what the higher tier adds, but what you do when three engines are not enough, because nothing on this page fixes that.

Which AI engines does it track?

Ten are available: ChatGPT, Gemini, Perplexity, Claude, Google AI Overviews, Google AI Mode, DeepSeek, Copilot, Grok and Mistral. You select three on Starter and Growth and five on Agency Growth, so the list is a menu rather than an inventory. Mistral appears on only one other tool we have reviewed.

Who is behind Superlines?

Grew Oy, based in Helsinki and founded in 2020, co-founded by Jere Meriluoto as CEO and Kimmo Ihanus as CTO. The site says it is backed by investors and a venture fund without naming amounts or firms, and that it participates in the NVIDIA Inception and Google Cloud AI Startup programmes.

Does Superlines publish how it measures?

It says it collects data from the real AI search interfaces rather than from APIs, which is the harder and better choice. It does not say how many times each prompt is run, what the sampling is, or how a mention is attributed to your brand. That is the same gap as every tool here except Evertune, and it is the question worth asking in the trial.

Is it a good fit for an agency?

The agency ladder is the part we would look at first. It runs at the same €79, €199 and €379 as the brand ladder and swaps brands for client brands, five of them on Agency Starter, and the partner programme adds white-label reporting, multi-client management, API and MCP access, and volume pricing that improves with client count. No partner figures are published, so the discount is a negotiation rather than a number.