Jenkins was the only CI/CD tool named in every answer we ran, a clean 6 of 6. On 2026-08-04 we put six ways of asking for the best CI/CD tool to ChatGPT and Gemini, gave each answer one live web search, and recorded every product named and every source domain behind it. This page is built from that measurement and nothing else.
Behind Jenkins, two tools tied at 5 of 6: CircleCI and GitHub Actions, each named in 83% of answers. GitLab CI/CD came next at 4 of 6 (67%). Then the drop is steep: Azure DevOps and Travis CI at 3 of 6 (50%) each, and everyone else at 2 of 6 or fewer. A buyer asking any of these six questions is very likely to hear Jenkins, CircleCI, and GitHub Actions before anything else.
How the shortlist changed from question to question
This is the part that should change how you think about being named. The six questions did not return one stable list. They returned overlapping but clearly different lists, from 3 names on the narrowest question to 7 on the widest, and the differences track the exact words in the query.
The bare "best ci/cd tools" question returned the widest list at 7 names, and it was the only question to name Codefresh (1 of 6, appearing nowhere else). The rest were the leaders: GitHub Actions, Jenkins, GitLab CI/CD, CircleCI, Travis CI, and Azure DevOps.
The "best free ci/cd tools" question changed the list the most. It named 6 tools, and three of them appear in this question and no other: Tekton, Drone, and the bare label "Travis," each 1 of 6. Just as striking, GitHub Actions and Azure DevOps both vanish from the free answer, even though GitHub Actions is named in 5 of the other answers. Ask about free, and the assistants swap hosted platforms for self-hosted open-source runners.
The "most affordable ci/cd tools" question returned the shortest list of all, just 3 names: GitHub Actions, GitLab CI, and Jenkins. Only Jenkins carries over from the free list. Free surfaced tools you host yourself (Tekton, Drone); affordable surfaced hosted tools with cheap tiers (GitHub Actions, GitLab CI). Same budget intent, two different shortlists.
The symmetry is worth a pause. The two 5-of-6 tools each miss exactly one question, and in both cases it is a budget one. CircleCI is named everywhere except "most affordable." GitHub Actions is named everywhere except "best free." Neither gap reads as a quality judgment; it is about which pages the search pulled for that exact phrasing.
Two more tools owe their entire presence to a single question. TeamCity (1 of 6) shows up only in "best ci/cd tools for small teams." Northflank and Harness (1 of 6 each) show up only in "ci/cd tools recommendations." Reword the question and those three names disappear from the results completely.
One label wrinkle, which we keep visible rather than smooth over: the same brand appears under two names. GitLab is "GitLab CI/CD" in 4 answers and the shorter "GitLab CI" in 2, and the short label lands in the budget-flavored questions (small teams and most affordable). Travis is "Travis CI" in 3 answers and bare "Travis" in the free one. We keep both labels on their own rows and do not add them into a combined total the measurement never showed.
The source map: which pages fed the answers
Three domains fed all six answers: reddit.com, blog.jetbrains.com, and orca.security, each cited in 6 of 6. That trio is the spine of every answer, one community forum and two vendor-run roundups.
Community threads carried a lot of the weight. Reddit was in 6 of 6, Quora in 4 of 6, and Stack Overflow, Facebook, and YouTube in 1 of 6 each. Roundup and vendor-blog pages did the rest: octopus.com in 4 of 6, atlassian.com and github.com in 3 of 6, then a tail of single-appearance sites (dev.to, thectoclub.com, lumenalta.com, incredibuild.com) at 1 each. Sixteen domains in total fed the six answers.
Classic review directories did not appear at all. Across all six answers there was no G2, no Capterra, no TrustRadius, no GetApp. For CI/CD tools, the assistants leaned on Reddit and vendor roundups, not the software-directory sites that many vendors pour budget into.
Here is the sharpest finding. A vendor's own domain fed an answer and its product was named in that same answer. In "ci/cd tools recommendations," northflank.com was one of the cited sources and Northflank was named (1 of 6). That same answer named Harness, and harness.io was a cited source in the measurement too. Northflank and Harness are not household CI/CD names, and they got a mention because their own published pages sat in the retrieval set. GitHub's own domain (github.com, 3 of 6) backs GitHub Actions the same way.
The counter-example matters just as much. atlassian.com was cited in 3 of 6 answers, yet no Atlassian CI/CD product (Bitbucket Pipelines or Bamboo) was named in any of the six. Getting your domain into the sources is not the same as getting your product named. Atlassian's pages fed the answers as roundups; the product still did not make a shortlist.
What we would do if we sold a CI/CD tool
If we sold a CI/CD tool, we would not try to out-market the category to get named; we would target these exact 16 source domains. Being in the Reddit and Quora threads, in the JetBrains, Orca, and Octopus roundups, and publishing one genuinely useful page on our own domain is what put Northflank and Harness into an answer at 1 of 6. The source list decides who gets named, and the source list is short and knowable.
Notice what raw size did not buy. Atlassian is far larger than Northflank, had its domain cited 3 times to Northflank's 2, and still did not get its product named in any of the 6 answers. So the work is specific: get onto the pages the assistants actually pull for your category, on the exact phrasings buyers use, because the "free" query and the "affordable" query pulled different pages.
Where Glotier fits, honestly
Glotier does not sell a CI/CD tool, so we are correctly absent from all 6 of these answers, and we are not going to pretend otherwise. But what you just read is the measurement, and it is exactly what we run for a customer's own category: the same six-question fan-out, the same named-product counts, the same source map that shows which pages to go win.
If CI/CD tools were your category, this is the report we would want in front of your marketing team every week: 14 products ranked, 16 source domains mapped, across 6 buyer questions. The check is free, no account, and about a minute to run for your own category.
The full list, counted
Counts are exactly as measured. GitLab and Travis each appear under two labels and stay on separate rows.
| Product | Named in | Share |
|---|---|---|
| Jenkins | 6 of 6 | 100% |
| CircleCI | 5 of 6 | 83% |
| GitHub Actions | 5 of 6 | 83% |
| GitLab CI/CD | 4 of 6 | 67% |
| Azure DevOps | 3 of 6 | 50% |
| Travis CI | 3 of 6 | 50% |
| GitLab CI | 2 of 6 | 33% |
| Codefresh | 1 of 6 | 17% |
| Drone | 1 of 6 | 17% |
Four more tools were each named in 1 of 6 (17%): Harness, Northflank, TeamCity, and Tekton, plus the bare "Travis" label noted above.