VOC AI (Shulex) Review 2026: Pricing, Report Caps and What It Actually Reads

VOC AI (Shulex) at a glance (2026)

Short answer: VOC AI — the product formerly and still widely called Shulex VOC — is an AI review-mining platform that clusters Amazon customer reviews into pain points and feature themes. As of data checked 2026-08-14, it is no longer sold on the cheap tiered plans most write-ups still quote. There is one personal paid plan, Pro at $99/month, a 3-day free trial with 2,000 credits, and two annual team plans at $599/year and $1,199/year. Everything is metered in credits, and the meter — not the plan name — is the thing to evaluate.

Buy it if you want fast thematic summaries of a competitor’s review corpus before you commit to a product. Skip it if you need a predictable monthly cost, responsive support, or a tool that keeps working on a bad day — those are the three complaints that dominate its public review record, and the newest of them is from October 2025, not the 2023 era most reviews cite.

One naming note that causes real confusion in search: the company brands the web app as VOC.AI, ships its Chrome extension as “ChatGPT for Amazon with GPT4 Shulex Copilot”, and still uses a shulex.com support address. All three are the same vendor.

VOC AI pricing: what the plans cost in 2026

The vendor publishes a single pricing page split into personal and team plans. These are the figures rendered in the page body on 2026-08-14:

Three things about this table are worth more than the numbers themselves.

There is no cheap monthly entry point any more. Many third-party write-ups still ranking for “shulex pricing” quote a $29 Basic and $69 Pro ladder; those tiers appear nowhere on the current page, so treat those sources as historical. The jump from a 3-day trial straight to $99/month is this product’s biggest commercial change, and it is why the trial deserves a real test plan.

The team plans are annual only, and the seat counts contradict the member counts. Team Lite is described for “3-5 members” but lists “5 team seats”; Team Growth is described for “10-20 members” but lists “15 team seats” (both quoted from the page, 2026-08-14). The vendor does not reconcile the two figures anywhere on the page, so confirm the seat number in writing before buying for a team of 20.

Both team plans render with a higher crossed-out “original” figure and no stated expiry, so the $599 and $1,199 numbers are what you can pay today.

For anyone who has read an AI-generated summary of this tool: the figures circulating as “$499/month” and “600 reports per month” do not match the published page. The real Pro report allowance is approximately 525 research reports per month, and $499 is not a price the vendor lists.

What the report and credit caps actually meter

Plan names tell you nothing here. The credit meter does, and it is fully reconstructible from the vendor’s own published approximations. Each plan lists a credit pool plus roughly how many “data queries” and “research reports” that buys. Divide one by the other and the same two constants fall out of every single plan:

Across all four, one research report costs 200 credits and one data query costs 80 credits (arithmetic derived from the vendor’s published figures, 2026-08-14; the vendor does not state these unit rates in words anywhere on the pricing page). That single fact makes the plans comparable in a way the marketing copy does not: a research report is 2.5 data queries, and nothing else on the page changes that exchange rate.

Run the consequences before you buy. A $99 Pro month spent entirely on research reports yields about 525 reports at roughly $0.19 each; Team Lite at $599/year yields about 4,200 at roughly $0.14 each — a 26% better unit rate, but it locks up a year to get there. If your real pattern is two or three product investigations a month, the Pro allowance is enormous overkill and you are paying for access, not volume.

The caps also matter for a non-budget reason. The vendor’s stated refund position, quoted by a buyer below, is that refunds are granted only when no reports have been generated. Under that rule, generating a single report inside the trial may forfeit your refund route — worth confirming in writing before you run anything.

What VOC AI actually reads

Review coverage is the part of this product that is least clearly documented, so here is what can and cannot be confirmed from the vendor’s own pages (checked 2026-08-14).

Confirmed on vendor pages. Amazon is the primary and most explicitly documented source — the homepage headline reads “VOC.AI - E-commerce Review AI Analysis for Sellers” and the analysis product page uses Amazon review parameters in its own interface examples. That page advertises “2B+ reviews” as the corpus scale and describes the core output as clustering “feedback by pain point, expectation, and feature mention”. The Chrome extension listing adds Shopify explicitly and claims it “works for all marketplaces”.

Named in vendor marketing but not in product documentation. Walmart, TikTok, Facebook and WhatsApp appear across the vendor’s site, but almost entirely in the context of its separate AI chatbot and customer-service product, not the review-analysis engine. If you are buying this to mine Walmart or TikTok Shop reviews, confirm that with the vendor rather than inferring it from a homepage logo strip.

Language coverage is two different things, and the site does not separate them. The web app’s interface offers English, Chinese, Japanese, German, Spanish, French and Portuguese; the extension listing separately claims 13 languages. Neither figure is stated as the number of buyer review languages the model can cluster, and the analysis page’s “buyer language” feature carries no supported-language list. That is a genuine documentation gap.

Worth stating plainly, because the category is muddled: VOC AI is an analysis layer. It does not send review requests and has no role in Amazon’s compliant solicitation flow — that is a separate class of tool covered in our Amazon review and feedback tools hub .

The Chrome extension is a separate product from the dashboard

Most people meet this vendor through the extension, and its numbers tell a different story than the web app’s. The listing — published by voc.ai, last updated 2026-08-06, with 60,000 users and a 4.5-star rating from 105 ratings (checked 2026-08-14) — sits an Amazon-page sidebar next to a listing and produces one-click review summaries, listing-copy suggestions, keyword ideas and competitor research.

Two structural points. First, 60,000 users against 105 ratings is a very thin feedback ratio, so 4.5 stars is a weak signal in either direction — neither validation nor grounds for reading the angry reviews as representative. Second, the extension has historically enforced its own free-tier limit that is not the web app’s credit meter: a 2023 reviewer hit a “daily limit of 10 analyses” in free extension mode, a cap the current pricing page does not mention. If you will work mainly in the sidebar, verify what your paid plan unlocks there rather than assuming the credit pool applies uniformly.

VOC AI on the 8-dimension scorecard

Scores follow the same 8-dimension methodology we apply to every tool on this site, weighted for a seller buying this specifically for review mining.

The scorecard splits cleanly in two. What the product does — cluster a large review corpus into readable themes, fast — scores well. What surrounds it — reliability, refunds, support and commercial terms — scores badly enough to decide the purchase for most buyers. That profile means the trial should be spent testing the boring parts, not the impressive ones.

What reviewers actually complain about

Four themes recur across the public record. All quotes below are verbatim from the platform indicated.

Output that cannot be traced to the product analysed. A 3-star Chrome Web Store review dated 2023-07-05 is the most specific: “It works great when it works, a lot of insights weren’t even comes from the product, and different products show the same identical reviews, not even mention that half of the time just won’t run even upgraded to Pro.” For a tool whose entire value is telling you what buyers said about this ASIN, identical output across different products is the failure mode that matters most. Spot-check any report against the raw reviews before you act on it.

Reliability under load. A 1-star Trustpilot review dated 2025-10-25 reads: “I signed up and paid $99 to run some reporting, and it didn’t work… multiple errors, slow performance… it works randomly, etc.” Note the $99 — that is a current-plan buyer, not a legacy-tier one, which makes this the most decision-relevant complaint in the set.

Trial and refund friction. A 1-star Trustpilot review dated 2025-10-07 describes being charged on a trial and refused a refund, quoting the vendor’s reply as: “We only refund if no reports were generated.” A separate 2-star review dated 2025-03-06 states the trial “required my card” and that three emails to [email protected] went unanswered. The current pricing page does not state whether the free trial requires a card (checked 2026-08-14), so this is unresolved rather than contradicted — check it at signup.

Free-mode limits that arrive faster than expected. A 1-star review dated 2023-09-08 describes hitting the free daily cap on the first click: “Currently in free mode, the daily limit of 10 analyses has been reached.” That reviewer’s upgrade prompt read “$99/month or $299/month” — the $99 has survived to today, a useful check that we are reading the same product.

One entry deserves explicit discounting: a 1-star review dated 2023-09-02 reads in full “All fake reviews, don’t trust this app” and contains no testable claim.

VOC AI vs the alternatives for Amazon review mining

Review mining is not a category with clean substitutes, because most tools bundle it into something larger. The honest comparison is by what you are actually buying access to (each vendor’s own published pricing; VOC AI checked 2026-08-14, Helium 10 checked 2026-07-20, DataHawk checked 2026-07-31, SageMailer checked 2026-08-13):

The comparison most buyers actually face is the last row. If you already pay for a general AI assistant, the gap VOC AI fills is collection and scale, not analysis — pulling and structuring thousands of reviews is the hard part, and summarising them is the part you can already do. That framing sets a clear test for the trial: if the reports read like something you could have produced yourself from a copy-paste of the first page of reviews, the tool is not earning $99.

Against Helium 10 the trade is breadth versus specialisation: if you need keyword and product research anyway, its suite price buys review tooling you will not separately budget for. DataHawk is barely comparable — an enterprise analytics warehouse with unpublished pricing, aimed at a different buyer. And none of these overlap with review solicitation tools like SageMailer or FeedbackWhiz , which generate reviews rather than read them.

Who VOC AI fits — and who should skip it

Good fit: private label sellers doing genuine pre-launch product research, where one well-clustered read of a competitor’s 3,000 reviews is worth more than $99; agencies that can spread a Team Lite pool across client work and absorb the annual commitment; and anyone who wants review themes without building a scraping pipeline.

Poor fit: sellers who need this a few times a year — the pricing has no shape for you since the cheap monthly tiers disappeared; teams of 15-20 who cannot get the seat count confirmed in writing; anyone buying primarily for Walmart or TikTok Shop review analysis, which is not documented on the review-analysis pages; and anyone whose workflow cannot absorb a tool that reviewers describe as working “randomly”.

Test during the 3 days, in this order: confirm whether the trial takes a card and what the refund position is before generating anything; run one report on a product whose reviews you already know and check the pain points are really in that corpus; then run a second on a different product and confirm the outputs differ — the specific failure reviewers describe, and five minutes to rule out.

FAQ

How much does VOC AI (Shulex) cost in 2026? Pro is $99/month, Team Lite is $599/year for 5 seats, and Team Growth is $1,199/year for 15 seats, plus a free 3-day trial with 2,000 credits (data checked 2026-08-14). The $29 and $69 tiers quoted in older articles are no longer published.

How many reports does a VOC AI plan include? Pro includes 105,000 credits a month, which the vendor states is approximately 525 research reports or 1,312 data queries. Working from the vendor’s own figures, a research report costs 200 credits and a data query 80.

Is VOC AI free? Only as a 3-day trial with 2,000 credits — roughly 10 research reports. There is no permanent free plan on the current pricing page. The Chrome extension has historically had its own separate free daily analysis cap.

Does VOC AI analyse Walmart and TikTok reviews? Amazon is the documented and primary source, with Shopify named on the extension listing. Other platforms appear mostly in the vendor’s separate chatbot and customer-service marketing, not on its review-analysis pages — confirm with the vendor before buying for those channels.

Related reading

• Amazon Review & Feedback Management Tools
• Amazon Review Checker Tools
• How to Remove Negative Customer Reviews on Amazon

Disclosure: this review contains no affiliate links. amzfinder does not accept paid placement or sponsored reviews. Scores reflect analysis of public vendor documentation and third-party review data, not a controlled lab benchmark. VOC AI pricing and product data checked 2026-08-14; competitor pricing dates are stated inline. Last updated: August 2026.