| Category | Function | Share | Rank | Negative rate | Labels | Quadrant | Since September 2026 |
|---|---|---|---|---|---|---|---|
| Data loss prevention | Security operations | 0% | 20 of 89 | 21% | 19 | accepted challenger | =heldSince September 2026: 0% → 0%, ±0 points. Inside the 11-point floor: within noise. Read over the models both editions asked. |
| DNS filtering | Network and edge | 0% | 77 of 94 | 50% | 6 | under 10 labels · led by DNSFilter at 59% | |
| SASE platforms | Network and edge | 0% | 40 of 45 | 100% | 1 | under 10 labels · led by Cato Networks at 35% | newNew since September 2026: not ranked then, 0% now. |
| Category | September 2026 | Now | Change | Reading | Rank |
|---|---|---|---|---|---|
| DLP | 0% | 0% | =heldSince September 2026: 0% → 0%, ±0 points. Inside the 11-point floor: within noise. Read over the models both editions asked. | Within noise | Rank 11 → 20 of 89 |
| DNS filtering | 0% | 0% | newNew since September 2026: not ranked then, 0% now. | New this edition | Rank 77 of 94, unchanged |
| SASE | 0% | 0% | newNew since September 2026: not ranked then, 0% now. | New this edition | Rank 40 of 45, unchanged |
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 | 0 | 1 | 1 | 0 | 2 |
| GPT-5.4 mini | 0 | 0 | 0 | 0 | 0 |
| Gemini 3.5 Flash | 0 | 1 | 0 | 0 | 1 |
| Perplexity Sonar | 0 | 1 | 0 | 1 | 2 |
| Grok 4.20 | 0 | 1 | 3 | 0 | 4 |
| Mistral Small | 0 | 1 | 1 | 0 | 2 |
| DeepSeek V4 Flash | 0 | 0 | 1 | 2 | 3 |
| Llama 4 Maverick | 0 | 1 | 1 | 0 | 2 |
| Qwen 3.7 Flash | 0 | 1 | 0 | 3 | 4 |
| Kimi K2 | 0 | 1 | 0 | 0 | 1 |
| GLM 4.7 FlashX | 0 | 0 | 0 | 0 | 0 |
| MiniMax M2.5 | 0 | 0 | 0 | 1 | 1 |
| GPT-6 Luna | 0 | 1 | 0 | 0 | 1 |
| Muse Glimmer 30B | 0 | 1 | 1 | 1 | 3 |
Verbatim evidence the judge attached to positive labels.
“If you want a Unified Platform (Security + DLP): Check Zscaler, Proofpoint, or Symantec” Qwen 3.7 Flash · DLP · scale prompt · alternative
“Zscaler Internet Access: Best for zero-trust DLP across cloud and internet traffic” Llama 4 Maverick · DLP · budget prompt · alternative
“Its strongest natural fit is DLP integrated with web/SSE and cloud-app controls” GPT-6 Luna · DLP · comparative prompt · alternative
“a solid option for mid-market B2B companies with distributed workforces” Mistral Small · DLP · direct prompt · alternative
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.
“Zscaler DLP and SecureTrust DLP ... described as having "limited classification" and "low visibility"” Perplexity Sonar · DLP · negative prompt · soft negative
“without the high price tag of enterprise suites like Cisco Umbrella or Zscaler” Qwen 3.7 Flash · DNS filtering · budget prompt · soft negative
“May be overkill if you only want DNS filtering; part of a larger suite.” Qwen 3.7 Flash · DNS filtering · scale prompt · soft negative
“Excellent performance, but may require higher minimums for discounts.” Qwen 3.7 Flash · SASE · budget 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 69 of the 77 answers that named Zscaler Internet Access and are not a share of its labels.
225 of the 225 domain citations in answers naming Zscaler Internet Access came from somebody else's page.
Pages are listed as the models cited them.
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.
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 Zscaler Internet Access's own pages state, read October 5, 2026: help.zscaler.com/zia/what-internet-saas, help.zscaler.com/zia/about-cloud-applications, help.zscaler.com/zia/about-internet-saas-users, help.zscaler.com/zia/managing-internet-saas-use-china, help.zscaler.com/zia/adding-internet-saas-admin-roles. 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 Zscaler Internet Access'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 Zscaler Internet Access, 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 Zscaler Internet Access by email, built from the raw record of the edition. It shows: