Which AI visibility tool a Singapore buyer should shortlist, sorted by what each vendor publishes
Four vendor pricing pages read on 27 October 2026, the one Singapore question none of them answers, and what two assistants named when the question was asked with Singapore in it.
Published by AI Knows Us · Updated 27 October 2026
Shortlist on one thing a buyer can check without a sales call: whether the vendor puts a figure and an engine count in the document a machine can read. Read on 27 October 2026, Otterly publishes three monthly figures against three prompt counts in its served page, Scrunch publishes a floor of 250 US dollars a month in its own page description, and Profound and Peec publish no figure at all on their pricing pages, routing instead to a sales conversation. For a Singapore buyer there is a second filter and it decides the invoice rather than the shortlist: which legal entity bills you, and whether Singapore GST at 9 per cent sits on it.
This page gives the readings with the page each one came from and the day it was read, the list of sources two assistants reached for when a Singapore buyer asked which tool to buy, and the four questions that make two quotes in this category comparable. Nothing here is a search volume, because no search volume exists for a question like this one.
The reading this page is built on
On 26 October 2026 we ran 64 blind pipelines across 32 batches on ChatGPT and Gemini with Singapore written into the text of every question, because the interface we query accepts a country parameter and answers the same way whether or not it is set. In the batch of six tool buying questions, the only AI visibility specialist named in the answer prose, on both engines, was Profound. Across all 32 blind batches our own name appeared zero times. The brand batch, which names us inside its own prompt, is excluded from every count on this page.
Being cited is a separate finding on a separate measure, and the two are never added. In that same audit our own domain was used as a source five times, and it was among the cited hosts in 3 of the 26 batches that carried a buyer question.
The sources two assistants cited when asked which tool to buy
That batch returned 20 distinct hosts. Four of them are the vendors a buyer would expect: peec.ai, otterly.ai, www.tryprofound.com and scrunch.com. The rest were www.measure.co, citations.io, sourcewatch.ai, anystat.ai, askwatch.ai, www.citationworks.com, suparank.io, klarvu.com, www.verifiedtools.info, www.generaldataworks.com, addure.sg, ai-cdn.scrunchai.com, www.imf.org, www.investing.com and data.gov.sg.
Read that list as a buyer rather than as a scoreboard. Most of it is made of names that a Singapore marketing lead has never heard, sitting beside two financial data sites and a Singapore government data portal. An answer assembled from that set is not a shortlist, it is a reading of whichever pages were available to quote. That is exactly why a published figure is worth more than a recommendation here: it is the one attribute you can verify yourself in under a minute.
What each vendor served on 27 October 2026
| Vendor and page read | What the served document carried | The unit the figure is attached to |
|---|---|---|
| Otterly, otterly.ai/pricing | Lite at 29 US dollars a month, Standard at 189 and Premium at 489, with a fourth tier shown as Custom | Search prompts, stated as 15, 100 and 400 |
| Scrunch, scrunchai.com/pricing | Pricing described as starting at 250 US dollars a month. The per tier figures were not in the served document | Monthly responses, site audits and brand workspaces, several shown as Custom |
| Profound, tryprofound.com/pricing | No figure. Both tiers on the page read Custom | Not published |
| Peec, peec.ai/pricing | No figure. Talk to Sales, with a stated 15 per cent discount for annual billing | Prompts, models and projects, with no number attached |
| Semrush, semrush.com/prices | The ladder renders after a script runs. Only a zero appeared in the served text, so the figures are unverified from that document on this date | Daily tracked prompts inside a wider platform |
| AI Knows Us, our own ladder | Published, in US dollars, with Growth at 199 a month outside the United States and the United Kingdom | Buyer questions and credits, which are two different units |
Two notes belong with that table. A vendor publishing no price is reported here as publishing no price, which is an answer to the buying question rather than a gap in this page. And no figure above is promised to be current. Each one is true of the page it was read from on 27 October 2026 and should be re-read before a purchase order is raised.
Where a rival beats us, on the same criterion
The criterion is what a vendor publishes that a buyer can check, so it has to be applied to us as well.
- Otterly attaches its figure to one unit. Fifteen, one hundred and four hundred search prompts against three prices is a comparison a buyer can do in their head. Our ladder counts buyer questions and credits separately, and a Claude check costs three credits rather than one, so working out what a month costs takes arithmetic rather than reading.
- Ahrefs bills a Singapore buyer from a Singapore company. Our own reading of that purchase, taken on 26 October 2026, records Ahrefs Pte. Ltd. as the contracting entity, which puts the subscription under Singapore law and Singapore GST. We cannot offer that, and for a finance team that prefers a local invoice it is a real advantage.
- Two vendors sell the URL that earned the citation and we do not. Our own ranking of this category for an in-house lead, read against vendor pages on 20 October 2026, found Ahrefs Brand Radar and Rankscale selling cited pages at URL level. We export the full tracking log instead, which is the raw material for that analysis rather than the view itself.
- We publish no Singapore dollar figure anywhere. Our own Singapore pricing page on why the invoice does not arrive in Singapore dollars sets out the whole column and states the gap. Budget the US dollar figure and let your finance team hold the conversion.
The four questions that make two quotes comparable
- How many engines does this tier read, as a number and a list. Five surfaces are in common use for this work: ChatGPT, Gemini, Perplexity, Claude and Google AI Overviews. A tier count is not an engine count, and on our own ladder two tiers read none at all.
- What is the unit, a prompt, a credit or a question. These are not interchangeable. A prompt allowance consumed once per engine, per location and per refresh is a different purchase from a credit balance.
- Which entity invoices, and is GST charged. IRAS states the current GST rate in Singapore as 9 per cent, read on iras.gov.sg on 27 October 2026. A foreign supplier of a digital service to a Singapore business sits in a different position from a local one, and the answer belongs in writing before the first renewal.
- Can I export the raw readings. A number you cannot export is a number you cannot audit, and in twelve months you will want to compare it with the reading somebody else took.
Whether a monthly subscription beats checking by hand
Checking by hand is the right first move and a poor second one, and the reason is measured rather than argued. Our own run to run test, published in full in what our own null result says about paying for this work, sent 22 buying questions to both engines three times each inside a six minute window on 22 October 2026. A brand that appeared at all appeared in all three runs 53.0 per cent of the time on ChatGPT and 26.1 per cent on Gemini. One reading of one prompt therefore cannot support a decision, and a person checking manually once a month is taking exactly one reading. What a tool buys is repetition on a fixed question list, with the answers kept, which is the only way a change next quarter is distinguishable from the ordinary wobble between two runs.
What an agency buying this on behalf of clients should check first
Three things decide it, and only one of them is the price. The first is how many brand workspaces a tier allows and whether a client can be given a login, because an agency that has to share one seat is running a reporting bottleneck. The second is whether the export carries the sources behind each answer, since a client will ask which page produced a mention. The third is the refresh cadence, because an agency reporting monthly needs the reading taken on a date it chooses rather than on the vendor's schedule. Our own pricing lists one login with full access on every plan, which is a genuine constraint for a team of three people and is stated here rather than written around.
Questions a Singapore buyer asks about AI visibility tools
Is there one tool that watches ChatGPT, Gemini, Perplexity and Claude at once
Yes, several, and the thing to verify is the engine list rather than the claim. Ask for the names in writing and ask what happens to the count when one of them changes its interface. Our own plans read those four plus Google AI Overviews, and the cheapest of our tiers that reads all five is Growth at 199 US dollars a month outside the United States and the United Kingdom.
What does this software cost per month for a Singapore company
On the pages we read on 27 October 2026 the published entry figure in the category was 29 US dollars a month at Otterly for 15 search prompts, and the published floor at Scrunch was 250 US dollars a month. Profound and Peec published no figure. None of these vendors quoted a Singapore dollar price, so every one of those numbers becomes a conversion on your own card statement.
Which tool did the assistants recommend when we asked
Profound, on both engines, in the batch of six tool questions we ran on 26 October 2026. That is a reading of what two assistants said on one day with Singapore in the prompt. It is not a verdict on the products, and the same batch cited a dozen hosts that no buyer in this market would recognise.
Are these tools worth paying for if my category is small
A small category is the case where the reading is cheapest to take and easiest to act on, because the number of buyer questions worth watching is small. Start with a free reading, write down the questions that matter, and buy a subscription at the point where somebody outside your team starts arguing with the number.
Why does this page print other vendors' prices
Because a buyer cannot compare what nobody will write down, and because a figure with a source and a date can be checked. Every number here names the page it came from and the day it was read, and any figure we could not verify from the served document is marked unverified on that date rather than filled in from memory.
The number worth writing on the purchase order
Three of the six vendors above will not tell you what they charge until you talk to somebody, and the shortlist the assistants handed us was mostly made of sites a Singapore buyer has never seen. That is the real state of this category on 27 October 2026, and it means the comparison you can actually run is a comparison of disclosure: engine count, unit, invoicing entity, export. AI Knows Us publishes its own ladder, its own method and its own failures, including the zero namings in the audit above, and the first reading at aiknowsus.com needs no account, so you can see what the five surfaces say about your category before any vendor on this page gets a purchase order.