Best QA Apps for Zendesk, Intercom and Gorgias (2026): 11 Marketplace Apps Scored

The 11 apps at a glance

Tiers: 8 (apps 1-3), 7 (apps 4-6), 5 (apps 7-9), 4 (app 10), 3 (app 11). Inside a tier, apps are ordered by Zendesk review count, then installs, then alphabetically, so the order isn't a quality ranking. Prices were read on each listing and vendor page on October 29, 2026.

How we scored them

We started from three pools. On Zendesk, the Marketplace has no QA category, so we took the 20 third-party apps our catalog query tags as QA, plus QA apps filed under WEM and other categories. Gorgias files 16 apps under Quality Assurance, several of them surveys, chargeback tools or AI agents. Intercom also has no QA category; its QA apps sit among 112 Analytics apps. We kept third-party apps whose main job is reviewing, scoring or enforcing the quality of agent conversations. Each app is scored 0 to 2 on five criteria.

Many apps list on more than one desk, so criterion 1 uses each app's best listing. Zendesk shows reviews; Intercom and Gorgias don't, so there we score badges and featured collections, then documentation. None of the 16 Gorgias QA cards shows a partner badge, and none of the ranked apps sits in an Intercom featured collection. Works in the ticket scores 2 only for an app in the ticket sidebar or editor that acts on the ticket; sidebars that only display a score get 1. An app with no Marketplace install gets 1 for setup. For maintenance we use the "Last updated" date a Zendesk listing shows, the dates in a Gorgias listing's page data, and the newest image upload on an Intercom listing.

Two apps here, evaluagent and AXIS by CraftCX, also appear in our Intercom analytics and QA roundup with the same five scores. They're here again because each has a second desk: evaluagent has a Zendesk sidebar app and CraftCX has a Gorgias listing.

Zendesk's own QA product sits outside the ranking. It began as Klaus, which Zendesk bought in February 2024, and it's the default most Zendesk teams compare against; our Zendesk QA guide covers its features and bundle price. Klaus still has a Gorgias listing, with "Pricing available upon request," page data last modified in January 2022, and a documentation link that returned a 404 when we checked.

Level AI and TheLoops are left out because our Intercom roundup already ranks them, and Scorebuddy and Playvox have no QA app on the Zendesk Marketplace or in Gorgias's QA category.

The 11 apps, ranked in tiers

Top tier, 8 of 10: MaestroQA, Guardrails and Score AI

1. Maestro QA Auto Quality Assurance: the review count, under a new name, 8/10

MaestroQA's Zendesk listing promises automated QA scorecards, "a side-by-side QA scorecard view" for grading tickets, and coaching built on AI-generated metrics. It has 69 reviews averaging 4.94, the second-largest review base in this list, and Zendesk shows it as last updated on April 9, 2026. Luca Messina wrote on the listing more than four years ago that "the support staff has been nothing short of phenomenal." There's no install button. The listing says to "contact our team" to get started, so setup scores 1.

Scores: reviews 2, price 2, works in the ticket 1, setup 1, maintained 2.

The catalog lists the app's organization as Rippit, and maestroqa.com/pricing now redirects to Rippit's pricing page, which lists a Free plan with $100 in AI credits, Starter at $185 a month, Growth at $495 and custom Enterprise. When we wrote our MaestroQA guide in September, the pricing page was a contact form. Rippit calls itself "The AI Conversation Data Platform" and doesn't mention MaestroQA, and its Intercom app, also called Rippit, signs up self-serve. Ask which plan covers the Zendesk scorecards before you budget on those numbers.

Best for: QA teams that already run a scorecard program on Zendesk and want the most-reviewed option.

2. Guardrails: QA before the ticket saves, 8/10

Guardrails checks an agent's update when they save it. Admins write rules the way they'd write a Zendesk trigger, with conditions on fields, groups, roles and assignee status, plus AI guidelines that review a reply against "brand, tone, empathy, and quality standards." When a check fails, the save is blocked and, in the listing's words, "agents receive a clear explanation of what went wrong and how to fix it." That makes it the one app here that acts inside the ticket.

Scores: reviews 0, price 2, works in the ticket 2, setup 2, maintained 2.

The listing shows three plans. Free covers 5 guardrails, 2 AI guidelines and 300 AI checks a month; Professional costs $5 per agent per month after a 14-day trial, with unlimited AI checks; Enterprise costs $9 and adds an audit log and a zero-retention mode. It was updated on October 9, 2026 and has no reviews yet. The maker, Aidvisor, describes itself as a Zendesk Tech Alliance Partner.

Best for: teams whose QA findings keep repeating, who'd rather stop the mistake than score it afterward.

3. Score AI: per-ticket scoring on Gorgias and Intercom, 8/10

Score AI pulls resolved tickets into Score and evaluates each against your evaluation forms, with coaching, sentiment and topic tracking, and roleplay training "based on actual performance data from their Gorgias interactions," per its Gorgias listing. Its Gorgias setup article is self-serve: enter your Gorgias domain in Score's integration settings and authorize the app. The Intercom guide is less open, telling you to contact your "assigned sales executive" if the integration appears disabled.

Scores: standing 1, price 2, works in the ticket 1, setup 2, maintained 2.

The pricing page lists Starter at $600 at $0.12 a ticket for up to 5,000 tickets a month, and a custom Scale plan from $0.08 a conversation above that. The Gorgias listing was last published in April 2026. Per-ticket pricing means the bill follows volume, not headcount, which suits a small team with busy queues and costs more for a team whose AI agent answers most of the tickets.

Best for: Gorgias and Intercom teams who want every ticket scored without a seat-based contract.

Second tier, 7 of 10: Kaizo, Pluno and AXIS by CraftCX

4. Kaizo: the most-reviewed QA app on Zendesk, 7/10

Kaizo offers AutoQA, AI conversation reviews, sentiment-based ticket sorting, performance drill-ins and gamified goals for agents. It has 104 reviews averaging 4.90 and 277 installs, the most of any QA app in the Zendesk catalog. The listing says it's "an out-of-the-box solution with a 2-min install," opened from Zendesk's left menu bar, and Zendesk shows it as updated on September 30, 2026. Charlie wrote on the listing over two years ago that it "greatly simplifies the QA team's work."

Scores: reviews 2, price 0, works in the ticket 1, setup 2, maintained 2.

Price costs it the top tier. Kaizo's pricing page lists Insight, Intelligence and Enterprise, all "Let's talk," billed per seat, where a seat is anyone who logs in, including the agents being reviewed. It says plainly, "We do not offer a self-serve trial." Our analytics and QA tools guide reviews Kaizo in more depth.

Best for: Zendesk teams that want a proven QA tool and are happy to start with a sales call.

5. AI Quality Assurance by Pluno: scores on the ticket after it closes, 7/10

Pluno's app scores every ticket and chat against criteria you write in plain text, as pass/fail or 1 to 5, with an explanation for each score. Agents see the evaluation "right in the Zendesk sidebar" once a ticket closes, and leaders get agent, criteria and ticket views in Pluno's dashboard. Setup runs through Pluno's onboarding page, which the listing says takes under 10 minutes. It was updated on October 15, 2026 and has 6 reviews, all five stars. A reviewer named Teberillo wrote: "This made QA less manual for us."

Scores: reviews 0, price 2, works in the ticket 1, setup 2, maintained 2.

The sidebar shows a score but doesn't change the ticket, so it gets 1 there. Pluno's pricing page lists AI Quality Assurance as an add-on at about $40 (€35) per agent per month, beside flat base plans that run from about $280 (€249) a month for up to 500 tickets to $5,150 (€4,649) for 30,000, with a 14-day trial of every module. The page doesn't say whether the QA add-on needs a base plan, so ask. Pluno also sells AI agents, so it has an incentive to show you which tickets score well enough to hand over.

Best for: teams that want agents to see their own scores the moment a ticket closes.

6. AXIS by CraftCX: a score for the AI agent, not the team, 7/10

CraftCX scores AI-led conversations on resolution accuracy, interaction effort and handoff smoothness. On Gorgias, its setup guide has you add the integration, approve the connection and run a backfill on recent tickets. The pricing page charges $0.05 per monitored conversation after the first 250, with no minimum and unlimited users. The Gorgias listing also mentions "Monthly subscriptions with included usage from $125/month," which we couldn't find on the pricing page.

Scores: standing 1, price 2, works in the ticket 1, setup 2, maintained 1.

It loses a point on maintenance: the Gorgias listing was published on October 1, 2025, four weeks outside our 12-month window, and the Intercom listing's newest images are from August 2025. Measuring an AI agent needs criteria human scorecards don't have, as we explain in our post on AI agent quality scoring, and AXIS ships them fixed.

Best for: teams that want one comparable number for their AI agent while human QA continues elsewhere.

Third tier, 5 of 10: evaluagent, EdgeTier and Intryc

7. evaluagent: a QA platform with a Zendesk window, 5/10

evaluagent is a full QA and coaching platform. Its Zendesk app does one thing, which its listing describes as letting you "view the evaluations completed on Zendesk tickets without leaving the Zendesk application." You install it by generating an API key and secret in evaluagent and pasting them into Zendesk. Its single review, from more than six years ago, reads: "Allows my users to view evaluation results on tickets that have evaluated within EvaluAgent." Zendesk shows the listing as updated on April 4, 2025.

Scores: standing 1, price 1, works in the ticket 1, setup 1, maintained 1.

The pricing page shows two seat-based plans, "from $35" and "from $65" per user per month, plus per-conversation pricing for AI agents on request. The pricing page lists no trial and offers a proof of concept after a demo, though the Zendesk listing still points to a free 30-day trial.

Best for: contact centers already on evaluagent that want agents to see results in Zendesk.

8. EdgeTier: Auto QA for large contact centers, 5/10

EdgeTier analyzes every ticket and chat for anomalies, sentiment and topics, and reviews conversations for quality. Its listing is unusually clear about data: it reads ticket content "including private/internal comments," stores it on AWS in Europe and deletes it after 3 months. You request credentials from EdgeTier before installing, and the app lives in Zendesk's top bar, not the ticket.

Scores: reviews 0, price 1, works in the ticket 1, setup 1, maintained 2.

EdgeTier's pricing page says plans start at about $22,300 (€20,000), priced on interactions analyzed rather than seats, and that it suits contact centers with more than 150,000 interactions a year. It was updated on September 15, 2026.

Best for: high-volume operations that want QA and incident alerts from one platform.

9. Intryc: QA and training for human and AI agents, 5/10

Intryc automates QA scorecards across "communication skills, processes, product knowledge, and operational actions," and supports review, dispute and calibration workflows, coaching and role-play simulations. Its listing says it was founded in 2024, is backed by Y Combinator and has raised more than $4 million. The Gorgias connection is read-only and ingests tickets up to 3 months back.

Scores: standing 1, price 0, works in the ticket 1, setup 1, maintained 2.

There's no price anywhere: the Gorgias listing says to book a demo, intryc.com/pricing returned a 404, and the homepage says, "Book a demo and our team will walk you through setup." The Gorgias listing was published in March 2026.

Best for: teams that want human and AI agents reviewed and trained in the same tool.

Fourth and fifth tiers: Loris (4) and 2clix (3)

10. Loris: on all three desks, with a vendor that has moved on, 4/10

Loris is the only vendor here with listings on Zendesk, Intercom and Gorgias. Its Zendesk app offers real-time coaching in the ticket sidebar and says it analyzes "100% of your Messages" for quality. The Gorgias listing says Loris "is priced based on agent seats and the number of channels covered" and to contact Loris.

Scores: standing 1, price 0, works in the ticket 1, setup 1, maintained 1.

loris.ai now redirects to Contentsquare's conversation intelligence page. The Zendesk listing was updated on April 4, 2025, and setup starts with a contact form. Confirm with Contentsquare that the help-desk apps are still sold before planning around them.

Best for: existing Loris customers checking what their help-desk app still does.

11. 2clix for Support: evaluations on the ticket, from Brazil, 3/10

2clix, a Brazilian company in the market "since 2012," per its listing, lets agents see a ticket's QA rating, evaluation status and key issues in the Zendesk sidebar, with a link to the full evaluation in its Quality Portal. New users start with a demo.

Scores: reviews 0, price 0, works in the ticket 1, setup 1, maintained 1.

We couldn't find a price. The listing's website, 2clix.com.br, redirects to 2clix.cx, which blocked our reader and has no archived copy.

Best for: Portuguese-speaking teams already auditing quality in 2clix.

Per seat or per conversation: what the published prices reward

At 10 agents, Guardrails Professional costs $50 a month and Pluno's add-on about $390 (€350). At 2,000 conversations a month, CraftCX costs $100, while Score AI lists its Starter plan at $600 for up to 5,000 tickets. The unit decides who pays for growth, and each vendor's business model shows in it. Seat pricing earns a vendor more as your team grows; conversation pricing earns it more as volume grows, including conversations an AI agent answers with nobody on the team touching them.

What QA apps change for an AI agent on the same tickets

When Macha runs AI agents in a help desk, the tickets these apps score are the ones our agents read and write. On Zendesk, Macha's tools read the ticket, its comments and custom field values, and write status, priority, custom fields, tags, internal notes and public replies.

We check four things when a QA app is already installed. First, whose replies it scores. Intryc reviews "Human and AI CX Agents" and CraftCX scores AI-led conversations, so decide whether the AI agent shares the team's scorecard or gets its own. Second, what it reads. EdgeTier reads private comments, so an agent's internal notes become QA data too. Third, what it blocks. Guardrails stops saves that fail its checks, so we ask the vendor whether those checks apply to updates that don't come from the agent workspace, and we read field definitions before an agent writes to a field another app owns. Fourth, what it tags. A QA app that adds tags can wake an agent's trigger, which is why Macha's triggers add a unique tag when they fire and exclude it from their own conditions.

Macha has its own agent evaluation, with automatic and manual scoring, and Studies that check an agent's performance against its instructions. Before anything runs on its own, we replay past tickets in a simulation and confirm whether replies post as internal notes or public.

Where AI fits

Every app above judges conversations, during or after they happen. An AI agent answers them, sets fields and tags, and hands off what it shouldn't handle.

Macha runs AI agents inside the help desk you already use, on Zendesk, Freshdesk, Gorgias, Front, HubSpot or Intercom, set up and monitored by the Macha team. It fits teams that want agents acting inside the ticket on routine requests while a QA app keeps grading the results. Pricing starts at $299/month for 750 tickets, with a trial of $50 of free usage (about 125 tickets), no credit card, no time limit. For the rest of a Gorgias stack, see our Gorgias integrations guide.

Which one fits your situation?

• You want the most-reviewed QA app on Zendesk: Kaizo, or MaestroQA if you're happy with a sales-led setup.
• The same mistakes keep showing up in QA: Guardrails, to stop them at save.
• You're on Gorgias and want every ticket scored: Score AI, or Intryc if you also want calibration and training.
• You only need to grade an AI agent: AXIS by CraftCX.
• Agents should see their own scores right after closing a ticket: Pluno.
• You run a large contact center with an annual QA budget: EdgeTier or evaluagent.
• You're on Zendesk and want the native option: Zendesk QA, outside this ranking.

FAQs

What are the best QA apps for Zendesk?

On our rubric, MaestroQA and Guardrails score 8 of 10 among Zendesk apps, and Kaizo and Pluno's AI Quality Assurance score 7. Kaizo has the most reviews, 104 averaging 4.90. Zendesk's own QA product, formerly Klaus, is the built-in alternative.

Which QA apps work with Gorgias?

Score AI, Intryc and CraftCX have current Gorgias listings with documentation. Loris and Klaus also list there, but Loris's domain now redirects to Contentsquare and Klaus's Gorgias page data was last modified in January 2022.

Can these QA apps score AI agent conversations?

Several say so. Intryc covers "Human and AI CX Agents," CraftCX is built for AI-led conversations, and evaluagent prices AI-agent conversations per conversation on request. Check whether AI replies land on the same scorecard as your team's.

<div class="docs-callout"><div><p class="docs-callout-title">Are you listed here?</p><p class="docs-callout-body">Partners and app makers can send corrections with a public source to [email protected]. We fix facts within a week. Scores follow the rubric above, not requests.</p></div></div>

How we researched this

On October 29, 2026 we queried three app stores: the Zendesk Marketplace catalog of 1,812 apps, the Gorgias App Store's Quality Assurance category of 16 apps, and the Intercom App Store's public data of 735 apps. We rendered every Zendesk and Gorgias listing in a browser the same day and read each vendor's pricing page and setup docs. Many Zendesk listings show the same "Last updated" date, April 4, 2025; we scored the date as shown. 2clix's site blocked us and had no archived copy. We didn't install or test any app or contact any vendor, and one reviewer applied the scores to public pages only.

Prices are shown in US dollars; prices published in other currencies were converted at European Central Bank reference rates of October 29, 2026.

Disclosure: Macha makes AI agents that run inside help desks. Nobody paid to be listed, and Macha is not ranked here.