Jaeger was the only distributed tracing tool named in every answer: 6 of 6, when we put six buyer questions about distributed tracing tools to ChatGPT and Gemini on 6 August 2026. No other tool ran the table. Four names tied right behind it at 5 of 6 answers each: Grafana Tempo, OpenObserve, SigNoz, and Zipkin. That is the entire top of the list, five open-source or open-source-first tools, before a single paid APM suite shows up.
We run this the same way every time. We put each of the six questions to ChatGPT and Gemini with one live web search, then we log every product named and every source domain the answer leaned on. The counts here are plain tallies of what came back across those six answers. Nothing is weighted and none of it is our opinion.
Below the leaders, Uptrace was named in 4 of 6 answers, then a three-way tie at 3 of 6: Honeycomb, OpenTelemetry, and SkyWalking. OpenTelemetry is the open instrumentation standard rather than a tracing backend of its own, and it still landed in 3 of 6 answers, a sign of how much these assistants frame this category around the open standard instead of a single product.
How the shortlist changed across the six questions
The most useful thing in this data is that there is no one shortlist. Only Jaeger appeared in all six questions; three tools appeared in just one; and the set turned over hard depending on how the buyer phrased the ask.
Ask the bare "best distributed tracing tools" and the answer was all open-source: Jaeger, OpenTelemetry, Zipkin, SkyWalking, Uptrace, SigNoz, Hypertrace, and Apache SkyWalking, eight names, and not one of them a paid SaaS suite. Neither Datadog nor New Relic appeared, and even Grafana Tempo and OpenObserve, both 5 of 6 overall, missed the generic "best" query entirely.
The paid names only surfaced when the buyer signalled budget or team size. Datadog APM and New Relic were each named in exactly 2 of 6 answers, and in both cases those two answers were the same pair of questions: "best distributed tracing tools for small teams" and "most affordable distributed tracing tools." Ask about small teams or affordability and the commercial vendors appear; ask anything else and they are gone.
Honeycomb followed that same budget cluster: 3 of 6 answers, and all three were the "small teams," "best free," and "most affordable" questions. It never showed up on the plain "best," "what should I use," or "recommendations" queries.
"Best free" and "most affordable" look like twins but are not. The free answer named seven tools: Uptrace, SigNoz, Jaeger, Zipkin, OpenObserve, Grafana Tempo, and Honeycomb. The affordable answer named eight: that same set minus Zipkin, plus Datadog APM and New Relic. Swapping one word, "free" for "affordable," was enough to drop Zipkin and pull two paid vendors in.
Three tools were named in only one of the six questions each. Hypertrace and Apache SkyWalking appeared only on the bare "best distributed tracing tools" query, at 1 of 6. Lumigo appeared only on "what distributed tracing tools should I use," also 1 of 6. A tool can be a legitimate answer to one phrasing and invisible to the other five, which is the whole reason tracking one query tells you almost nothing.
The source map: who fed the answers
Two domains fed every single answer: openobserve.ai and reddit.com, each cited in 6 of 6 questions. One is a vendor blog and the other is a community forum, and together they were the spine of this category's answers.
Reddit was the only community forum in the whole source list, cited in all 6 answers. There was no Quora and no Stack Overflow. The crowd signal in distributed tracing, as these assistants read it, comes from Reddit and nowhere else.
The key insight is that vendors who publish get named. OpenObserve's own domain fed all 6 answers and OpenObserve the product was named in 5 of 6. Uptrace's own domain fed 4 answers and Uptrace was named in 4. The same held for Jaeger (its site cited in 2 answers, the product named in 6), New Relic (site in 3, product in 2), and Lumigo (site in 2, product in 1). Owning a page the assistant reads is not a guarantee, but every vendor whose domain fed these answers also got its product named.
Not every vendor domain paid off. Chronosphere's roundups fed 4 of 6 answers and Turbo360's fed 4, yet neither Chronosphere nor Turbo360 was named as a tool in any answer. Dash0 (cited in 3), Last9 (2), Coralogix (1), Groundcover (1), and Sematext (1) told the same story: their content was read, their own product was not recommended. Writing the roundup gets you cited; it does not get you listed.
Review directories were absent. Across all 16 source domains behind these answers, G2 and Capterra did not appear once. In plenty of B2B categories those two dominate; in distributed tracing, on this run, they contributed nothing, and vendor docs, Reddit, and independent roundups such as turbo360.com, nurbak.com, and medium.com did all of the work.
What it takes to get named in this category
If you sell a distributed tracing tool and want to be in these answers, the source list is your to-do list, not your market share. The assistants named whoever the pages in front of them named, and there were only 16 source domains in play.
From this data, concretely: be on Reddit, which was in all 6 answers, and keep a technical page of your own that reads like real documentation, since openobserve.ai and uptrace.dev show a vendor page can carry a product into 4 to 6 answers. Get into the recurring independent roundups, because turbo360.com fed 4 answers and dash0.com fed 3. And write for intent, since Datadog APM and New Relic only surfaced on "affordable" and "small teams," a page aimed at one specific buyer situation can win a query that the generic "best" list will never hand you.
The full list, counted
| Product | Named in | Share |
|---|---|---|
| Jaeger | 6 of 6 | 100% |
| Grafana Tempo | 5 of 6 | 83% |
| OpenObserve | 5 of 6 | 83% |
| SigNoz | 5 of 6 | 83% |
| Zipkin | 5 of 6 | 83% |
| Uptrace | 4 of 6 | 67% |
| Honeycomb | 3 of 6 | 50% |
| OpenTelemetry | 3 of 6 | 50% |
| SkyWalking | 3 of 6 | 50% |
| Apache SkyWalking | 1 of 6 | 17% |
Two of those rows are the same project. "SkyWalking" at 3 of 6 and "Apache SkyWalking" at 1 of 6 are one tool the assistants labelled two ways, so we list both exactly as they were returned instead of summing them into a "4 of 6" the measurement never actually produced. Just off the table, Datadog APM and New Relic were each named in 2 of 6 answers (33%), and Hypertrace and Lumigo in 1 of 6 (17%).
Where Glotier fits
Glotier does not sell a distributed tracing tool, so we are correctly absent from all 6 of these answers, and we would not want to fake our way in. This is us doing our own homework on someone else's category, in public.
But this page is exactly what we do for a customer's own category. We take the six or so ways your buyers really ask, run them through the same assistants, and hand back the named-count table and the source map for your space, so you can see whether you are the Jaeger of your category, named every time, or the Chronosphere, read constantly but never named. The check is free, needs no account, and takes about a minute.