Kyriba was the single most-recommended product in this category, named in 4 of the 6 answers we got when we asked ChatGPT and Gemini to pick the best treasury management software. No other product matched it.
We ran six buyer questions past ChatGPT and Gemini on 7 August 2026, one live web search behind each answer, and logged every product named and every source domain cited. Kyriba led at 4 of 6. Three products tied right behind it at 3 of 6 each: GTreasury, Nomentia, and Rho.
The six questions were the ones a real buyer types: best treasury management software, best for small teams, what should I use, best free, most affordable, and a plain request for recommendations. Each got one answer, so every count below is out of 6.
Below the top four, six products were named in exactly 2 of the 6 answers: Coupa Treasury, ION Treasury, Ripple Treasury, SAP Treasury and Risk Management, TIS, and TreasuryView. A long tail of ten more products showed up in just 1 answer each, which means 20 distinct product names came back across only six questions.
The shortlist changed with the question
The wording of the question moved the shortlist more than anything else did. The plain "best treasury management software" query returned 8 names, and the "best free treasury management software" query returned only 3, with almost no overlap between the two.
The bare "best" answer was an enterprise lineup of 8: Kyriba, GTreasury, Nomentia, TIS, SAP Treasury and Risk Management, ION, HighRadius Treasury, and Rho. The "recommendations" query returned 6 names and introduced nothing new, all repeats of that same enterprise core.
Only four names were consistent across the phrasings: Kyriba, Nomentia, Rho, and GTreasury each appeared in 3 or more of the 6 answers. Every other product depended on exactly how the buyer worded the question.
"Best free" returned just 3 names, and two of them appear nowhere else in the entire run: Agicap and Atlar. A buyer who types the word "free" meets two products that a buyer who types "best" never sees. TreasuryView was the third free name, and it also surfaced for small teams.
"Most affordable" returned a different 3: Nomentia, Rho, and Coupa. Here Coupa appeared under its bare name for the only time in the run, distinct from the "Coupa Treasury" label that showed up in 2 other answers. So the affordable question was the only place the plain "Coupa" name came back.
The two remaining questions each added their own names. "For small teams" returned 8 and was the only answer to name Trovata, DebtBook, and HighRadius. "What should I use" returned 7 and was the only answer to name Cobase and FIS Treasury and Risk Manager. In total, ten products were named in only one of the six questions: Agicap, Atlar, Cobase, Coupa, DebtBook, FIS Treasury and Risk Manager, HighRadius, HighRadius Treasury, ION, and Trovata.
Where the answers came from
Four domains sat behind all six answers, cited in 6 of 6: gartner.com, nomentia.com, rho.co, and treasury.ripple.com. Three of those four are vendors' own websites, and that is the finding that matters most.
Every vendor whose own domain fed the answers also got named, and the correlation is tight. Nomentia's site was cited in 6 of 6 answers and Nomentia was named in 3. Rho's site fed 6 of 6 and Rho was named in 3. Ripple's treasury.ripple.com fed 6 of 6 and Ripple Treasury was named in 2. Kyriba's own domain fed 4 of 6 answers, and Kyriba was the most-named product of all at 4 of 6. The assistants were largely naming the vendors whose pages they were reading.
Community threads barely registered. Reddit appeared in 1 of 6 answers, on the bare "best" query only. LinkedIn appeared in 1 of 6. Quora did not appear at all. For a category this technical, there was almost no community-forum signal in the sources.
The analyst and directory picture was lopsided. Gartner was cited in all 6 answers, the joint-highest source. The peer-review directories that vendors spend on were nearly absent: G2 appeared in exactly 1 of 6 answers, on the "best free" query, and Capterra did not appear at all. Roundup and vendor blogs filled the rest, with debtbook.com cited in 4 of 6 answers and both highradius.com and treasuryview.com in 3 of 6 each.
What gets a treasury vendor named
The source list, not market share, decides who gets named across all six answers. The assistants named products they found on the pages that ranked for these queries, and a vendor absent from those pages was absent from the answers regardless of its real install base.
The clearest lever is your own domain. Four vendors were cited from their own sites in 6 of 6 or 4 of 6 answers, and all four were named. If your pages rank for "best treasury management software" and its variants, you enter the pool the assistants read from. If they do not, you are invisible to this buyer no matter how large you are.
Gartner is the second lever, cited in 6 of 6 answers. The peer-review directories are a weak lever in this category: with G2 at 1 of 6 and Capterra at 0 of 6, effort spent there moved almost none of these answers. And because "free" and "affordable" pulled almost entirely separate names, winning the price-conscious buyer means being visible on the specific pages behind those two questions, not only the "best" roundups.
The full list, counted
| Product | Named in | Share |
|---|---|---|
| Kyriba | 4 of 6 | 67% |
| GTreasury | 3 of 6 | 50% |
| Nomentia | 3 of 6 | 50% |
| Rho | 3 of 6 | 50% |
| Coupa Treasury | 2 of 6 | 33% |
| ION Treasury | 2 of 6 | 33% |
| Ripple Treasury | 2 of 6 | 33% |
| SAP Treasury and Risk Management | 2 of 6 | 33% |
| TIS | 2 of 6 | 33% |
| TreasuryView | 2 of 6 | 33% |
Three vendors came back under two different labels, and we kept each label separate rather than inventing a combined total the data never showed. "Coupa Treasury" was named in 2 answers and "Coupa" in 1. "ION Treasury" in 2 and "ION" in 1. "HighRadius Treasury" in 1 and "HighRadius" in 1. Counting by the exact label each assistant used is the honest way to report what actually came back, so we did not fold these into single higher numbers.
Where Glotier fits
Glotier does not sell treasury management software, so we are not named in any of these six answers, and that is exactly right. We would be worried if a treasury buyer's assistant recommended us for a category we are not in.
What you just read is the thing Glotier does, pointed at someone else's category. We ask ChatGPT and Gemini the questions your buyers actually type, we count who gets named across the six answers, and we show you the source pages behind each one, the same gartner.com and vendor domains you saw above. If you sell treasury software, the six-answer version of this for your own product runs in about a minute, with no account.