| Category | Function | Share | Rank | Negative rate | Labels | Quadrant | Since September 2026 |
|---|---|---|---|---|---|---|---|
| API gateways | Developer platform | 11% | 2 of 66 | 8% | 26 | accepted challenger | ▲+7Since September 2026: 4% → 11%, +7 points. Inside the 11-point floor: within noise. Read over the models both editions asked. |
| LLM gateways and observability | Data platform | 2% | 10 of 69 | 0% | 2 | under 10 labels · led by LiteLLM at 31% | ▲+2Since September 2026: 0% → 2%, +2 points. Inside the 11-point floor: within noise. Read over the models both editions asked. |
| API management | Developer platform | 0% | 17 of 61 | 12% | 8 | under 10 labels · led by Kong Gateway at 37% | ▼−2Since September 2026: 2% → 0%, −2 points. Inside the 11-point floor: within noise. Read over the models both editions asked. |
| Category | September 2026 | Now | Change | Reading | Rank |
|---|---|---|---|---|---|
| API gateways | 4% | 11% | ▲+7Since September 2026: 4% → 11%, +7 points. Inside the 11-point floor: within noise. Read over the models both editions asked. | Within noise | Rank 5 → 2 of 66 |
| API management | 2% | 0% | ▼−2Since September 2026: 2% → 0%, −2 points. Inside the 11-point floor: within noise. Read over the models both editions asked. | Within noise | Rank 9 → 17 of 61 |
| LLM gateways | 0% | 2% | ▲+2Since September 2026: 0% → 2%, +2 points. Inside the 11-point floor: within noise. Read over the models both editions asked. | Within noise | Rank 35 → 10 of 69 |
Shares here are read over the models both editions asked, so they can differ by a point or two from the standing above, which counts every model in this edition.
The floor is 11 points of share, measured: how far the models move a leader on their own when the same questions are asked twice with nothing changed. A larger change is movement; a smaller one is noise, and both are shown. Movement is read over the twelve models both editions asked; GPT-6 Luna, Muse Glimmer 30B joined this edition and are in the standing but not yet in the comparison. How the floor is measured · The editions
| Model | First choice | Alternative | Mention | Negative | Labels |
|---|---|---|---|---|---|
| Claude Haiku 4.5 | 1 | 0 | 0 | 0 | 1 |
| GPT-5.4 mini | 1 | 0 | 0 | 0 | 1 |
| Gemini 3.5 Flash | 2 | 3 | 1 | 0 | 6 |
| Perplexity Sonar | 0 | 2 | 0 | 0 | 2 |
| Grok 4.20 | 1 | 2 | 0 | 2 | 5 |
| Mistral Small | 0 | 0 | 1 | 0 | 1 |
| DeepSeek V4 Flash | 0 | 2 | 3 | 0 | 5 |
| Llama 4 Maverick | 1 | 0 | 0 | 0 | 1 |
| Qwen 3.7 Flash | 1 | 1 | 0 | 0 | 2 |
| Kimi K2 | 1 | 2 | 0 | 0 | 3 |
| GLM 4.7 FlashX | 1 | 0 | 2 | 1 | 4 |
| MiniMax M2.5 | 0 | 0 | 1 | 0 | 1 |
| GPT-6 Luna | 0 | 0 | 0 | 0 | 0 |
| Muse Glimmer 30B | 1 | 2 | 1 | 0 | 4 |
Verbatim evidence the judge attached to positive labels.
“Use Kong (for its vast ecosystem) or Apache APISIX (if you want pure open-source with no relational DB requirements)” Gemini 3.5 Flash · API gateways · comparative prompt · first choice
“Apache APISIX and Traefik are currently the strongest recommendations for budget-conscious teams” Qwen 3.7 Flash · API gateways · budget prompt · first choice
“Go with Kong OSS or Apache APISIX - you get enterprise-grade features at zero license cost” Kimi K2 · API gateways · budget prompt · first choice
“Apache APISIX consistently outperforms other open-source gateways in published benchmarks” Claude Haiku 4.5 · API gateways · comparative prompt · first choice
Verbatim evidence attached to negative labels. A warning on a product with few labels is a warning; on a product with many, it is one voice among them.
“Be cautious. These platforms are powerful, but they require a highly skilled team to secure properly.” GLM 4.7 FlashX · API management · negative prompt · soft negative
“Emerging, smaller community than Kong” Grok 4.20 · API gateways · paraphrase prompt · soft negative
“Similar misconfig exposure as Kong” Grok 4.20 · API gateways · negative prompt · soft negative
Citations exist only for the models that return a source list, five of the fourteen in this edition, so these counts come from 75 of the 81 answers that named Apache APISIX and are not a share of its labels.
262 of the 262 domain citations in answers naming Apache APISIX came from somebody else's page.
Pages are listed as the models cited them.
First shown March 2025, as GridGain Systems, verified. 1 more by other advertisers linking to it.
On Google's pages: 1
Search figures are US estimates from DataForSEO, read September 28, 2026; AI search demand is its modeled, directional estimate, not a count of queries to any assistant. The answers are this edition's. Two measurements side by side: neither is read as the cause of the other.
| Kind | Pages | Last 90 days | 2025-10 to 2026-09 | Latest | Categories named |
|---|---|---|---|---|---|
| News or press | 1,081 | 12 | 2026-09-03 | ||
| Blog | 228 | 1 | 2026-07-01 | ||
| Template or tool | 38 | undated | |||
| Report or ebook | 14 | undated | |||
| Conference or event | 10 | 0 | 2026-06-09 | ||
| Glossary or explainer | 9 | undated | |||
| Case study | 2 | undated | |||
| Webinar or virtual event | 2 | undated |
Every page apache.org exposes, subdomains included. Kind is read from the address and title. The last 90 days, the latest date and the twelve months count pages by when they were published, from the site's feeds, a date in the address, or the page's own publication date, read from up to a hundred of its most recently changed pages; a page that says only when it last changed is counted in its kind but not in when, so the recent counts are a floor, and a kind none of whose pages gives a publication date reads undated. An event counts as online when its address or title says so (webinar, on demand, virtual or online summit); a conference, summit, trade show, expo or roadshow that does not say so is counted as a conference or event, which on a vendor's site is mostly in person. Read September 29, 2026.
An email the morning each edition publishes: where this product moved, where it held, and by how much against the noise floor. One address, confirmed by a click; a stop link in every email.
Already following? Everything you follow, with a stop for each.
What Apache APISIX's own pages state, read October 5, 2026: apisix.apache.org, apache.org, apache.org/use-cases/dish. A claimed page can correct any of them.
Claiming is free and changes nothing in the data. A claimed page shows a verified contact who is told when each edition publishes and when Apache APISIX's standing changes by more than the noise floor; the right to propose corrections to the vendor table, meaning names the judge wrote that should or should not read as Apache APISIX, applied by version and listed in the change log; and a one-line description supplied by the vendor and marked as such.
A new claim receives the current edition's vendor brief for Apache APISIX by email, built from the raw record of the edition. It shows:
Apache APISIX is among the products AI models recommend first in API gateways this edition. Each badge says so in the buyer's words, names the edition, and links to the standing. The next edition issues a new badge; this one stays true as a record of October 2026.
Badges read on light and dark pages. Alt text carries the claim, the category, the edition and the index, so it stays a citation where the image does not load.