| Category | Function | Share | Rank | Negative rate | Labels | Quadrant | Since September 2026 |
|---|---|---|---|---|---|---|---|
| Web application firewalls | Network and edge | 69% | 1 of 82 | 7% | 67 | endorsed leader | ▼−21Since September 2026: 89% → 67%, −21 points. Past the 11-point floor: movement. Read over the models both editions asked. |
| Category | September 2026 | Now | Change | Reading | Rank |
|---|---|---|---|---|---|
| WAF | 89% | 67% | ▼−21Since September 2026: 89% → 67%, −21 points. Past the 11-point floor: movement. Read over the models both editions asked. | Moved past the floor | Rank 1 of 82, 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 | 3 | 1 | 0 | 0 | 4 |
| GPT-5.4 mini | 2 | 1 | 0 | 0 | 3 |
| Gemini 3.5 Flash | 4 | 2 | 0 | 0 | 6 |
| Perplexity Sonar | 2 | 1 | 1 | 0 | 4 |
| Grok 4.20 | 3 | 1 | 1 | 1 | 6 |
| Mistral Small | 3 | 2 | 1 | 0 | 6 |
| DeepSeek V4 Flash | 5 | 0 | 0 | 1 | 6 |
| Llama 4 Maverick | 1 | 0 | 1 | 0 | 2 |
| Qwen 3.7 Flash | 5 | 0 | 1 | 0 | 6 |
| Kimi K2 | 3 | 1 | 1 | 0 | 5 |
| GLM 4.7 FlashX | 4 | 0 | 0 | 1 | 5 |
| MiniMax M2.5 | 3 | 0 | 1 | 0 | 4 |
| GPT-6 Luna | 3 | 0 | 1 | 1 | 5 |
| Muse Glimmer 30B | 4 | 0 | 0 | 1 | 5 |
Verbatim evidence the judge attached to positive labels.
“Cloudflare WAF (Best All-Rounder & Cloud-First B2B)... widely considered the most accessible and comprehensive WAF for small-to-mid-market businesses” Gemini 3.5 Flash · WAF · direct prompt · first choice
“Massive global scale, edge programmability, OWASP Core Rule Set-based rules, and easy self-serve setup. Consistently named a Forrester Wave Leader.” DeepSeek V4 Flash · WAF · comparative prompt · first choice
“1. Cloudflare WAF - Strengths: Global threat intelligence, machine learning, and a massive CDN for low latency and high performance.” Mistral Small · WAF · comparative prompt · first choice
“the recommended WAF and bot protection service for a mid-sized B2B company is Cloudflare. It is considered the best overall and best value” Llama 4 Maverick · WAF · paraphrase 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.
“Misconfigured "Big Three" WAFs (Cloudflare, Akamai, Imperva)... you have effectively bypassed the WAF. ... Do not avoid these vendors, but avoid the configuration” GLM 4.7 FlashX · WAF · negative prompt · soft negative
“Cloudflare's managed-rule body limit varies by plan, and content beyond it may not be fully analyzed.” GPT-6 Luna · WAF · negative prompt · soft negative
“Free tiers of popular cloud WAFs (e.g., Cloudflare Free) offer very limited protection” DeepSeek V4 Flash · WAF · negative prompt · soft negative
“parsing discrepancies affect AWS, Azure, Cloud Armor, Cloudflare and ModSecurity alike” Muse Glimmer 30B · WAF · 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 178 of the 203 answers that named Cloudflare WAF and are not a share of its labels.
614 of the 614 domain citations in answers naming Cloudflare WAF 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 Cloudflare WAF's own pages state, read September 28, 2026: blog.cloudflare.com/waf-ml, developers.cloudflare.com/waf, blog.cloudflare.com/account-waf, blog.cloudflare.com/new-cloudflare-waf, blog.cloudflare.com/waf-rules-react-vulnerability. 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 Cloudflare WAF'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 Cloudflare WAF, 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 Cloudflare WAF by email, built from the raw record of the edition. It shows:
AI models recommend Cloudflare WAF first in Web application firewalls 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.