> ## Content Index > Fetch the complete content index at: https://www.theaidirectory.com.au/llms.txt > Use this file to discover other available public pages before exploring further. # Trunk Review Australia 2026: Green CI, US-Only Data - URL: https://www.theaidirectory.com.au/review-trunk-io-australia-2026/ - Published: 2026-06-29T23:00:00.000Z - Updated: 2026-10-09T06:37:04.000Z - Description: A genuinely useful flaky-test and merge-queue platform with a real free tier — held back for Australians by US-only data residency and no local billing. - Author: Adam - Tags: Coding & Development, AI Coding & Developer Tools, #review **16**/25 ★★★★☆ Very Good Trunk Verified 9 Oct 2026 Reviewed 6 Sep 2026 Value for Money**4**/5 Ease of Use**3**/5 Feature Depth**4**/5 Australian Fit**3**/5 Reliability & Support**2**/5 Pricing Pricing on application — see the pricing table **The gate:** Five stars need 21+ and Australian Fit of at least 4/5. [Skip to the verdict ](#verdict) > **Verdict (TL;DR):** Trunk is a CI reliability platform, not an AI coding agent. It detects and quarantines flaky tests and runs a smarter GitHub merge queue. The free tier is genuinely useful, and the Team tier is also priced at zero per committer. Its only AI is failure analysis that summarises broken builds. For Australians, the catch is location: data residency is US-only and billing is US-denominated with no local entity. A strong tool for an engineering team that lives on GitHub, but it does not pretend to serve a non-technical Australian. > > **Score: 16/25 — ⭐⭐⭐⭐ Very Good.** > > *Last reviewed: 6 September 2026 by Adam.* ## What it is Trunk is a platform for keeping continuous-integration pipelines green. It is built for engineering teams, not solo non-technical users. The vendor describes it as "the CI reliability platform trusted by teams that ship fast." It does two main jobs. First, it finds flaky tests — tests that pass and fail at random — across any language or CI provider, then quarantines them so they stop blocking your pipeline. Second, it layers extra intelligence onto GitHub's merge queue: failed pull requests stay in the queue while others keep testing, batches of up to 100 pull requests are grouped and auto-bisected when they fail, and hotfixes get their own lane. There is a third, separate product. Trunk Code Quality is a meta-linter: a single tool that downloads, installs, and runs dozens of third-party linters, formatters, and security scanners across a polyglot repository. The vendor's documentation describes it as "a metalinter and static analysis manager." It scans only changed code, runs a background daemon for speed, and annotates inside VS Code and Neovim. Where is the AI? Narrower than the category name suggests. Trunk markets "AI-powered failure analysis" — it spots duplicate test failures and writes a plain summary of what went wrong. That is useful triage. It is not an AI coding agent that writes or reviews your code. Buyers expecting a Copilot-style assistant should look elsewhere. ## Pricing in AUD Trunk's pricing is unusual: the first two tiers both read "$0/committer/month." The model bills on usage — test spans — not seats. The **Free** tier costs nothing and needs no credit card. It covers up to five repository committers and five million test spans per month, plus flaky-test quarantining, pull-request comments, the CLI and VS Code integration, and GitHub inline annotations. For a small team, that is a real working tier, not a teaser. The **Team** tier is also priced at US$0 per committer per month, but lifts the cap to unlimited committers and adds one million test spans per committer. On top of the Free features it adds the full merge queue with multi-CI support, parallelism controls, Slack notifications, onboarding support, and admin and authentication controls. There is a free trial of the Team tier with technical support. A note of caution on those zeros. Trunk's published pricing page lists no per-span overage rate for the Team tier; Enterprise is described as getting a "volume discount on test spans." How spend behaves once a high-volume team exceeds the included spans is not stated on the pricing page, so treat "$0" as the entry price, not a guarantee of a zero invoice at scale. Confirm overage terms with sales before you commit a large pipeline. The **Enterprise** tier is custom-priced, with custom payment terms, SAML, on-premise deployment, Slack Connect support, and volume discounts. There is no AUD pricing. Figures are US-denominated. Because the entry tiers are US$0, there is no headline conversion to run — but any Enterprise contract will be quoted in US dollars, and Australian GST treatment will depend on how Trunk's US entity invoices you. Budget for that conversation rather than assuming a clean local invoice. ## Australian fit This is where Trunk's score is held down, and the reason it cannot reach five stars under our methodology. Trunk's own pricing page is explicit about location: data residency is **US**, on both the Free and Team tiers. Data is encrypted in transit and at rest, and the platform is SOC 2 Type 2 compliant — both good signals. But "US data residency" means your test metadata and code-quality telemetry are processed and held in the United States, with no Australian or Asia-Pacific region offered on the standard tiers. For a lot of Australian engineering teams, that is acceptable. CI test metadata is lower-sensitivity than, say, customer records, and SOC 2 Type 2 is a credible control standard. But it is a firm "no" for any team under data-sovereignty obligations — government, defence, health, or regulated finance — that must keep code and build data onshore. On-premise deployment exists, but only on the custom Enterprise tier, which puts the onshore option behind a sales conversation and an unpublished price. On commercials, there is no Australian entity, no AUD billing, and no documented local support presence. Support is Slack-based and, at the top tier, Slack Connect. That is fine for an engineering audience comfortable in US-timezone, US-dollar tooling. It is a poor fit for our core reader — a non-technical Melbourne business owner — who is simply not the customer here. So the Australian-fit picture is honest but middling: low-sensitivity data, solid compliance, but US-only residency, US billing, and no local presence. That earns a 3, not a 4. ## Where it's strong The free tier is the headline strength. Most CI-tooling vendors gate the useful features behind a paid plan; Trunk gives flaky-test quarantining, pull-request comments, and IDE integration away at zero cost for up to five committers. A small Australian dev shop can adopt the core product without a purchase order. The flaky-test handling is the genuinely valuable bit. Flaky tests are a real, expensive problem — they erode trust in the suite and waste CI minutes on reruns. Trunk detects them across any language and runner, quarantines them automatically, and keeps a status history so you can see which tests are unreliable over time. The AI failure summaries make triage faster by grouping duplicate failures. The merge queue is the other strong card, and it is well-designed. Anti-flake protection, intelligent batching of up to 100 pull requests with auto-bisection, and parallel lanes for hotfixes are exactly the features a high-throughput team needs to stop the main branch breaking. For a team already shipping dozens of pull requests a day on GitHub, this is a real productivity lever. Trunk Code Quality is a useful free-standing bonus: one command to manage many linters across a mixed-language repository, scanning only what changed. ## Where it's weak The first weakness is the category mismatch. Filed under AI Coding and Developer Tools, Trunk is not an AI coding assistant. Its AI is limited to failure analysis. If you arrived wanting code generation, autocomplete, or AI code review, this is the wrong product — and the marketing's "AI-powered" framing can mislead. The second is GitHub gravity. The merge queue is built on top of GitHub's, and the deepest integrations assume GitHub. Teams on GitLab, Bitbucket, or Azure DevOps get less from the platform. This is a GitHub-first tool. The third, for Australians specifically, is location. US-only data residency on the standard tiers, US-denominated billing, no local entity, and onshore deployment locked behind Enterprise. None of that is disqualifying for a typical commercial team, but it rules Trunk out for sovereignty-bound workloads and adds friction for everyone else. Finally, the pricing transparency gap. Two "$0" tiers are attractive, but the absence of a published per-span overage rate means a large team cannot model its true cost from the website alone. You have to ask. ## The 5×5 score | Criterion | Score | Why | | --------------------- | --------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------- | | Value for Money | 4/5 | A genuinely useful free tier and a US$0 Team tier are rare in this category. Held off 5 only because high-volume span overage pricing is not published. | | Ease of Use | 3/5 | Clean for an engineering team on GitHub with a CLI and VS Code, but it is a developer tool — no value for a non-technical user, and best features assume GitHub. | | Feature Depth | 4/5 | Strong flaky-test detection, a well-designed merge queue, and a free meta-linter. Marked down because the "AI" is narrow — failure analysis, not coding. | | Australian Fit | 3/5 | Low-sensitivity data and SOC 2 Type 2 help, but US-only data residency, US billing, no local entity, and onshore deployment only on Enterprise. | | Reliability & Support | 2/5 | SOC 2 Type 2 and encryption are solid, but support is Slack-based with no documented Australian presence, and overage behaviour at scale is undocumented. | | **Total** | **16/25** | **⭐⭐⭐⭐ Very Good** | > **Australian Fit gate:** Trunk's raw total of 16/25 sits in the four-star band on its own, so the gate does not change its displayed rating here. But note the gate's other arm: with Australian Fit at 3/5, Trunk **cannot be named a category Top Pick** for AI Coding & Developer Tools, regardless of any future score gains, until its local-fit story improves. See our [Methodology](https://www.theaidirectory.com.au/methodology/) page. ## Verdict Trunk does one thing well and says so plainly: it keeps continuous integration green. The flaky-test quarantining is genuinely valuable, the merge queue is smartly built, and the free tier lets a small Australian team adopt the core product without spending a cent. For a GitHub-based engineering team drowning in flaky tests and merge bottlenecks, it earns its place. Two honest qualifiers keep it from a higher mark. It is not the AI coding assistant the category name implies — the AI is confined to summarising failures, and buyers should not expect more. And for Australians, the location story is firmly American: US-only data residency on the standard tiers, US-denominated billing, and no local entity or support presence. That is fine for most commercial teams and a hard no for anyone bound to keep build data onshore. **Buy it if** you run a fast-shipping team on GitHub, flaky tests and merge queues are costing you real time, and US data residency is acceptable for your CI metadata. **Look elsewhere if** you are under data-sovereignty obligations, you are not on GitHub, or you came looking for an AI coding assistant — Trunk is not that. Either way, start on the free tier; it tells you most of what you need before any sales call. Update history - **9 Oct 2026** — Re-checked for the AUD-claim queue (#447), trunk.io/pricing read on an Australian connection: Free and Team are still US$0 per committer a month, Enterprise is still quoted ("volume discount", "Contact Sales"), and no per-span overage rate is published — nothing changes. `aud_native: false` set in the frontmatter, which had no value. Capture: `Dossiers/captures/trunk-io-pricing-2026-10-09.txt`. - **6 Sep 2026** — Pricing re-verified live (Free and Team still US$0/committer/month) and Wayback-captured; the 5 Sep "not vendor-verified" flag removed; the sources block moved into the house Sources format. - **30 Jun 2026** — First reviewed. ## Sources - **Trunk — Pricing** (primary): [trunk.io/pricing](https://trunk.io/pricing?ref=theaidirectory.com.au) — [Wayback 30 June 2026](https://web.archive.org/web/20260630033746/https://trunk.io/pricing) · [Wayback 6 September 2026](https://web.archive.org/web/20260906054207/https://trunk.io/pricing). Re-verified live 6 September 2026: Free **US$0/committer/month** (up to 5 committers, 5M test spans/month); Team **US$0/committer/month** (unlimited committers, 1M test spans per committer); Enterprise custom with SAML, on-premise deployment and volume discounts. Unchanged since June. - **Trunk — Home / product overview** (primary): [trunk.io](https://trunk.io/?ref=theaidirectory.com.au) — [Wayback 30 June 2026](https://web.archive.org/web/20260630033804/https://trunk.io/). - **Trunk — Flaky Tests** (primary): [trunk.io/flaky-tests](https://trunk.io/flaky-tests?ref=theaidirectory.com.au) — [Wayback 30 June 2026](https://web.archive.org/web/20260630002504/https://trunk.io/flaky-tests). - **Trunk — Merge Queue** (primary): [trunk.io/merge-queue](https://trunk.io/merge-queue?ref=theaidirectory.com.au) — [Wayback 30 June 2026](https://web.archive.org/web/20260630002531/https://trunk.io/merge-queue). - **Trunk — Code Quality documentation** (primary): [docs.trunk.io/code-quality](https://docs.trunk.io/code-quality?ref=theaidirectory.com.au) — [Wayback 12 May 2025](http://web.archive.org/web/20250512165901/https://docs.trunk.io/code-quality). - AUD conversions in this review are indicative only — current rate, shown ex-GST — see [Methodology](https://www.theaidirectory.com.au/methodology/). *Trunk Code Quality (the meta-linter) is described in the vendor's documentation at docs.trunk.io/code-quality; a same-day Wayback capture could not be confirmed on the review date, so its details are attributed to that documentation rather than treated as independently snapshotted. Pricing is US-denominated; the published page lists no per-span overage rate, and Australian GST treatment of any Enterprise contract depends on how Trunk's US entity invoices the customer. Exchange rate reference: US$1 ≈ A$1.40 (September 2026).* ---