AI security platforms for enterprise buyers
Asked as “AI security platform”, and as “tool to secure LLM applications and employee AI use”, on behalf of an enterprise B2B company. 54 first choices recorded across the direct, paraphrase, budget and scale prompts, fourteen models each. Added to the October 2026 Edition on October 4, 2026; its answers are read by the distilled judge (ai-indexes-judge-qwen3-14b-run3), not the claude-opus-5 judge of the earlier categories: how the two compare.
Standing · first-choice share
9%
Contested · Check Point AI Security 7%
09SentinelOne Singularity07Check Point AI Security06CrowdStrike Falcon78others
9% of first choices, contested.
By buyer segment
The same question asked on behalf of a different buyer. Each standing is computed within its segment; they sit side by side and are never added together.
The standing
Share is the count of first choices across the direct, paraphrase, budget and scale prompts, over all fourteen models, for an enterprise B2B company. Ordered by share.
ProductFirst-choice shareNegative rateLabelsQuadrant
Show the four products at 0%, ordered by negative rate
Bars are the share of first choices, 0 to 100Every product with at least 10 labels here. Every product name links to its product page.
For buyers
Recommended versus criticized
Every product with at least 10 labels here, on both axes. The 30% line names a quadrant, not the verdict above: that one needs more than 40%.
Criticized challengerCriticized default
Negative label rate →
01
02
03
04
05
06
07
08Accepted challengerEndorsed leader
0%First-choice share → · lines at 30% share and 25% negative20%
Key
01SentinelOne Singularity9%
SentinelOne Singularity9%02CrowdStrike Falcon6%
CrowdStrike Falcon6%03Lakera Guard4%
Lakera Guard4%04Palo Alto Networks2%
Palo Alto Networks2%05HiddenLayer0%
HiddenLayer0%06Cyera0%
Cyera0%07CrowdStrike0%
CrowdStrike0%08Darktrace0%
Darktrace0%What they warned about
One of fourteen models held their first choice under the paraphrase. Claude Haiku 4.5, GPT-5.4 mini, Perplexity Sonar, Grok 4.1 Fast, Mistral Small, DeepSeek V4 Flash, Llama 4 Maverick, Qwen 3.7 Flash, Kimi K2, GLM 4.7 FlashX, MiniMax M2.5, GPT-6 Luna and Muse Glimmer 30B changed. A high negative share on a product with few labels is a warning. A low share on a product with many labels is salience, not sentiment.
Darktrace43%
6 of 14 labels negative · 5 of 14 models · 3 hard negative
“Avoid for Cost Predictability: Platforms like Darktrace (quote-based enterprise licensing, typically starting in six figures annually)” Claude Haiku 4.5, budget prompt
Prophet Security100%
5 of 5 labels negative · 5 of 14 models · 2 hard negative
“Custom quotes, tiered by investigation volume — costs scale with activity, not user count.” DeepSeek V4 Flash, budget prompt
Microsoft Sentinel50%
4 of 8 labels negative · 4 of 14 models · 3 hard negative
“**Avoid Usage-Based Models:** Steer clear of pure API-driven tools (like some iterations of Microsoft Sentinel or Wiz)” Qwen 3.7 Flash, budget prompt
Vectra AI50%
4 of 8 labels negative · 4 of 14 models · 1 hard negative
“the following platforms are not considered suitable for large enterprises or should be used with caution: 1. Vectra AI” Llama 4 Maverick, negative prompt
What they cite
Citations exist only for the models that return a source list: fourteen of the fourteen in this edition, and all six flagship models on the expanded tier.
Sites the answers cite
79 of 84 answers in this category came back with a source list, from 14 of 14 models: citations where the model returns them, or the search results it consulted. 1200 links across 377 sites, every framing counted. Ranked by the number of answers carrying the site or page. None of the 252 answers across every segment cited this index's own page for the category.
vendor site · Check Point34 answers · 46 citations · 11 models
vendor site · Cyera30 answers · 47 citations · 10 models
vendor site · Witness.ai23 answers · 39 citations · 11 models
vendor site · UnderDefense20 answers · 31 citations · 11 models
vendor site · Venn20 answers · 23 citations · 9 models
vendor site · Reco20 answers · 21 citations · 9 models
vendor site · Guideflow20 answers · 20 citations · 10 models
vendor site · SentinelOne19 answers · 33 citations · 8 models
vendor site · Palo Alto Networks18 answers · 37 citations · 9 models
16 answers · 20 citations · 7 models
vendor site · CrowdStrike12 answers · 21 citations · 8 models
vendor site · Nudge Security12 answers · 14 citations · 7 models
Pages the answers cite
The ten pages named in the most answers, by full address. A page here is one the models returned with a recommendation, not one the index endorses.
26 answers · 9 models
20 answers · 10 models
18 answers · 8 models
17 answers · 9 models
16 answers · 8 models
15 answers · 8 models
14 answers · 7 models
12 answers · 8 models
11 answers · 8 models
11 answers · 6 models
Search against answers
Each company's standing in the answers beside its site's footprint in Google search, one row a site: the products the models named on it with their shares, and the share they add up to; monthly searches on Google, and DataForSEO's estimate of AI search demand (modeled from search signals, directional, not a count of queries to any assistant), for the most-searched of the company's and its products' names (the name is in each row's hover text); estimated monthly organic visits to the site; and its best position in Google's top ten for “ai security platform”, “ai security platforms”, “best ai security platform”. US estimates from DataForSEO and Google's Ads Transparency Center. A small company's site, or a mid-sized company's site for its flagship, is marked company; a product on a large parent's site (Google, Microsoft) has no site figures. A column with no figures for this category is left out, and an empty cell means none were seen, not none exist. Two measurements side by side: neither is read as the cause of the other.
| Company and products | Share | Own site cited | Name searches, Google | AI search demand, est. | Organic visits | Paid search a month, est. | Google ads, last 30 days | Best Google position |
|---|---|---|---|---|---|---|---|---|
| SentinelOne sentinelone.com SentinelOne Singularity 9% | 9% | 534 | 33,100 | 1,675 | 446,702 company | $780 | 22 | |
| Palo Alto Networks parent site Palo Alto Networks Prisma AIRS 6%, Palo Alto Networks Cortex XSIAM 2%, Palo Alto Networks Cortex Cloud 2% | 9% | 227 | 40,500 | 1,088 | ||||
| Check Point parent site Check Point AI Security 7% | 7% | 13 | 14,800 | 4,487 | ||||
| CrowdStrike parent site CrowdStrike Falcon 6% | 6% | 234 | 74,000 | 5,130 | ||||
| Prompt Security prompt.security | 6% | 23 | 1,000 | 41 | 2,008 | |||
| WitnessAI witness.ai WitnessAI 4%, Witness.ai 2% | 6% | 13 | 590 | 14 | 3,200 | 1 | 3 | |
| Lakera Guard lakera.ai | 4% | 43 | 170 | 3 | 2,639 | |||
| Lasso Security | 4% | 590 | 3 | |||||
| Netskope netskope.com | 4% | 41 | 18,100 | 830 | 42,119 | 27 | ||
| Nudge Security nudgesecurity.com | 4% | 117 | 880 | 1 | 18,419 | 28 |
Google's top ten for the three searches hold 18 sites; 3 of them are among the sites the models cited here (cyera.com, paloaltonetworks.com, witness.ai). The first result for “ai security platform” is reddit.com.
How they answered
Six framings of the same buying question, each sent to every model in a fresh session with search on. One row per model, so a row shows whether it held its answer under rewording, what it named when cost was the constraint, and what it argued against. Computed from the raw judge labels.ShowHide
How they answered
Six framings of the same buying question, each sent to every model in a fresh session with search on. One row per model, so a row shows whether it held its answer under rewording, what it named when cost was the constraint, and what it argued against. Computed from the raw judge labels.
ShowHideModelDirect“What is the best AI security platform for an enterprise B2B company?”Paraphrase“Which tool to secure LLM applications and employee AI use would you recommend to a large B2B company with thousands of employees?”Comparative“What are the top enterprise-grade AI security platforms and how do they differ?”Budget-constrained“What is the best AI security platform for a large company that needs predictable total cost across thousands of users?”Scale-constrained“We are a 5,000 person company with SSO, SOC 2 and procurement review requirements evaluating an AI security platform. What should we look at?”Negative“Which AI security platforms should a large enterprise avoid or be cautious about?”
Claude Haiku 4.5BifrostWitnessAIChangedTrueFoundryNudge Securityagainst: Darktrace, Palo Alto Networks Cortex XDRno first choicenothing named
Four alternatives
CyberArk, Lakera Guard, Palo Alto Networks, TrueFoundryThree alternatives
Aim Security, Credo AI, RecoThree alternatives
HiddenLayer, Lasso Security, Palo Alto Prisma AIRSTwo alternatives
Google Cloud Chronicle, SentinelOne SingularityGPT-5.4 miniPalo Alto NetworksZscaler AI Security / AI ProtectChangedMicrosoft Defender XDRagainst: AWSCheck Pointagainst: CrowdStrikeno first choicenothing named
Three alternatives
CrowdStrike, Proofpoint, ZscalerTwo alternatives
Cloudflare AI Security for Apps, Netskope AI Security / AI GuardrailsThree alternatives
CrowdStrike: Falcon AI security capabilities, Google Cloud: Security Command Center AI Protection, Palo Alto Networks: Prisma AIRS / AI Gateway / AI Runtime SecurityOne alternative
NetskopeGemini 3.5 FlashPrompt SecurityPrompt SecurityHeldPalo Alto NetworksCato AI Security, Netskopeno first choicenothing named
Six alternatives
Cyera, Harmonic Security, Lakera Guard, NeuralTrust, Orca Security, WizFour alternatives
CalypsoAI Moderator, Lakera Guard, Netskope, Palo Alto NetworksFour alternatives
Check Point Software, Cisco AI Defense, HiddenLayer, SentinelOne SingularityThree alternatives
LiteLLM, TrueFoundry AI Gateway, Zscaler Secure AI AccessPerplexity SonarPalo Alto Networks Prisma AIRSCheck Point AI SecurityChangedagainst: Lakera LLM Guardno first choiceForcepoint Data Security Cloudno first choiceagainst: ChatGPT, Claude, Google Gemini, Microsoft Copilot, Moonshot AI's Kimi
Five alternatives
Cyera, Forcepoint Data Security Cloud, Microsoft’s AI security stack, Proofpoint, VennTwo alternatives
Lasso Security, Prompt SecurityOne alternative
Palo Alto Networks Cortex XSIAMGrok 4.1 FastCrowdStrike Falcon, SentinelOne SingularitySentinelOne's Prompt SecurityChangedagainst: GarakHiddenLayerSentinelOne Singularityagainst: Darktrace, Microsoft Sentinel, Palo Alto Networks Cortex XDRLasso Securityagainst: AWS, Check Point, CrowdStrike, Fortinet, Google Cloud, Google reCAPTCHA, Palo Alto Networks, Prophet Security, SentinelOne Singularity, Vectra AI, Wiz, hCaptcha
Four alternatives
Check Point, Darktrace, Palo Alto Networks, ZscalerThree alternatives
Check Point AI Security, Lakera Guard, Lasso SecurityFive alternatives
Cisco AI Defense, CrowdStrike Falcon AI, Cyera, Lakera Guard, Netskope One AI SecurityTwo alternatives
CrowdStrike Falcon, UnderDefense MDRThree alternatives
Cyera, Fortinet AI tools, Witness.aiMistral SmallCyberhaven, SentinelOne SingularityCheck Point AI Security, Prompt SecurityChangedagainst: GarakHiddenLayerCrowdStrike Falcon, SentinelOne Singularityno first choiceagainst: Prophet Security, TrueFoundry, Vectra AI
Three alternatives
Check Point Software, Fortinet FortiGuard, HiddenLayer AISec PlatformOne alternative
RecoThree alternatives
Check Point Software, CrowdStrike, Lakera GuardOne alternative
Dropzone AIDeepSeek V4 FlashPalo Alto Networks Prisma AIRSLakera Guard, Zscaler AI Access SecurityChangedPalo Alto Networks Prisma AIRSCrowdStrike Falconagainst: BigID, Cyera, Darktrace, Microsoft Sentinel, Prophet Security, Splunk Security Cloudno first choicenothing named
Seven alternatives
CalypsoAI, Credo AI, CrowdStrike Falcon, HiddenLayer, Lakera Guard, SentinelOne Singularity, WitnessAISix alternatives
Bifrost, Garak, Mindgard, Netskope, Reco, WizEight alternatives
Check Point — Lakera Guard, Cisco — Secure AI / Robust Intelligence, F5 — CalypsoAI, HiddenLayer, Microsoft — Defender for AI / Microsoft Purview, Mindgard, SentinelOne Prompt Security, Wiz AI-SPMFive alternatives
Microsoft Defender for Cloud, Microsoft Defender for Endpoint, SentinelOne Singularity, UnderDefense AI SOC, ZscalerLlama 4 MaverickCheck Point SoftwareImperva AI Application Security, Lakera Guard, Lasso SecurityChangedno first choiceno first choiceno first choiceagainst: Prophet Security, TrueFoundry, Vectra AI
Four alternatives
Cyera, Forcepoint Data Security Cloud, Nudge Security, VennQwen 3.7 FlashSentinelOne SingularityNetskope, ZscalerChangedProtect AIUnderDefense AI SOCagainst: CrowdStrike, Darktrace, Microsoft Sentinel, Wizno first choiceagainst: Lacework, Prisma Cloud, Wiz
One alternative
Witness.aiThree alternatives
Censinet, Lakera Guard, Microsoft PurviewSeven alternatives
AWS, Google Cloud, HiddenLayer, Lakera Guard, Microsoft, Mindgard, Robust IntelligenceKimi K2Palo Alto Networks Cortex CloudWitnessAIChangedCrowdStrike FalconMicrosoft Defender for Endpointagainst: CrowdStrike Falcon, Darktrace, Microsoft Sentinel, Palo Alto Networks Cortex XDRno first choiceagainst: CrowdStrike, Darktrace, SentinelOne Singularity
Three alternatives
CrowdStrike Falcon, Microsoft Security Copilot + Sentinel, SentinelOne SingularityThree alternatives
LangProtect, Noma Security, Prompt SecuritySix alternatives
Darktrace, IBM QRadar, Microsoft Sentinel, Palo Alto Networks Cortex Cloud, SentinelOne Singularity, Vectra AIOne alternative
SentinelOne SingularityGLM 4.7 FlashXPrisma AIRSCheck Point AI Security, Witness.aiChangedno first choiceMicrosoft Security Copilotno first choiceagainst: Hugging Face
Four alternatives
CyberArk Identity Security Platform, SentinelOne Singularity, WitnessAI, ZenityFour alternatives
BigID, HiddenLayer, Prisma AIRS, VennOne alternative
CrowdStrike Charlotte AIMiniMax M2.5Google Security Operations, Microsoft Copilot for SecurityMicrosoft 365 Copilot with PurviewChangedno first choiceMicrosoft Defender for Enterpriseagainst: Palo Alto Cortex, SentinelOne Singularityno first choiceagainst: Google Vertex AI
Three alternatives
AWS AI Security, CrowdStrike Falcon, SentinelOne SingularityFour alternatives
Cyberhaven, Reco, SentinelOne Prompt Security, WitnessAIOne alternative
CrowdStrikeGPT-6 LunaPalo Alto Networks Prisma AIRSCheck Point AI SecurityChangedPalo Alto Networks Prisma AIRSMicrosoft Purviewno first choiceagainst: Lakera Guard, Prompt Security, Protect AI
One alternative
Wiz AI-APPTwo alternatives
Netskope One, Palo Alto Prisma AIRSFive alternatives
Check Point, Cisco AI Defense, Lakera Guard, Microsoft’s AI security stack — Defender, Purview, Entra, Foundry, Agent 365, Wiz AI Security Posture ManagementTwo alternatives
Netskope, ZscalerMuse Glimmer 30BPalo Alto Networks Cortex XSIAM, SentinelOne Singularity Platform with Purple AIPrompt Security — now part of SentinelOne SingularityChangedWiz AI-SPMNudge Securityno first choiceagainst: DeepSeek, Prophet Security, Vectra AI
One alternative
Cisco AI DefenseOne alternative
WitnessAI Unified AI Security and Governance PlatformSeven alternatives
Blue Border, Check Point, Cyera, HiddenLayer, Lakera Guard, Mindgard, Palo Alto Networks Prisma AIRSOne alternative
Dropzone AIBold is the first choiceAlternatives are counted; the count opens them.What the answer argued against
The record
One row per call: the version string exactly as returned, whether the model searched, sources cited, and latency. Full answer text is in the free responses file. Download the record
Eighty-four rows: every prompt, every model, every answer.
| Prompt | Model | Version string | Time (UTC) | Searched | Sources | Latency |
|---|---|---|---|---|---|---|
| Direct recommendation | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-05 01:03 | yes | 9 | 8 s |
| Direct recommendation | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-05 01:21 | yes | 5 | 7 s |
| Direct recommendation | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-04 23:39 | yes | 19 | 25 s |
| Direct recommendation | Perplexity Sonar | sonar | 2026-10-05 00:57 | yes | 18 | 3 s |
| Direct recommendation | Grok 4.1 Fast | spacexai/grok-4.1-fast-non-reasoning via vertex | 2026-10-04 23:31 | yes | 24 | 8 s |
| Direct recommendation | Mistral Small | mistral/mistral-small via mistral | 2026-10-04 22:56 | yes | 14 | 8 s |
| Direct recommendation | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-05 00:27 | yes | 24 | 47 s |
| Direct recommendation | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-04 22:38 | yes | 5 | 2 s |
| Direct recommendation | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-04 22:56 | yes | 5 | 21 s |
| Direct recommendation | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-05 01:30 | yes | 20 | 23 s |
| Direct recommendation | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-05 01:18 | yes | 23 | 22 s |
| Direct recommendation | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-05 01:24 | yes | 5 | 18 s |
| Direct recommendation | GPT-6 Luna | gpt-6-luna | 2026-10-04 22:11 | yes | 3 | 18 s |
| Direct recommendation | Muse Glimmer 30B | meta/muse-glimmer-30b via togetherai | 2026-10-04 22:38 | yes | 25 | 23 s |
| Paraphrase | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-04 22:31 | yes | 18 | 10 s |
| Paraphrase | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-04 22:20 | yes | 3 | 5 s |
| Paraphrase | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-05 00:17 | yes | 18 | 25 s |
| Paraphrase | Perplexity Sonar | sonar | 2026-10-04 22:29 | yes | 18 | 4 s |
| Paraphrase | Grok 4.1 Fast | spacexai/grok-4.1-fast-non-reasoning via vertex | 2026-10-05 01:44 | yes | 21 | 7 s |
| Paraphrase | Mistral Small | mistral/mistral-small via mistral | 2026-10-05 00:23 | yes | 5 | 5 s |
| Paraphrase | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-05 01:33 | yes | 25 | 30 s |
| Paraphrase | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-05 01:31 | yes | 5 | 2 s |
| Paraphrase | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-05 00:41 | no | 0 | 25 s |
| Paraphrase | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-04 22:55 | yes | 15 | 15 s |
| Paraphrase | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-05 01:03 | yes | 24 | 41 s |
| Paraphrase | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-05 00:27 | yes | 15 | 35 s |
| Paraphrase | GPT-6 Luna | gpt-6-luna | 2026-10-05 00:07 | yes | 3 | 19 s |
| Paraphrase | Muse Glimmer 30B | meta/muse-glimmer-30b via togetherai | 2026-10-05 01:55 | yes | 22 | 30 s |
| Comparative | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-04 22:09 | yes | 9 | 9 s |
| Comparative | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-04 23:56 | yes | 10 | 11 s |
| Comparative | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-05 00:59 | yes | 29 | 31 s |
| Comparative | Perplexity Sonar | sonar | 2026-10-05 01:35 | yes | 20 | 5 s |
| Comparative | Grok 4.1 Fast | spacexai/grok-4.1-fast-non-reasoning via vertex | 2026-10-05 01:20 | yes | 21 | 10 s |
| Comparative | Mistral Small | mistral/mistral-small via mistral | 2026-10-04 23:49 | yes | 25 | 11 s |
| Comparative | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-05 00:42 | yes | 24 | 37 s |
| Comparative | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-05 01:38 | yes | 5 | 2 s |
| Comparative | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-05 01:06 | yes | 10 | 31 s |
| Comparative | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-05 01:32 | yes | 19 | 24 s |
| Comparative | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-04 22:53 | yes | 22 | 53 s |
| Comparative | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-05 01:17 | yes | 17 | 29 s |
| Comparative | GPT-6 Luna | gpt-6-luna | 2026-10-04 23:59 | yes | 7 | 23 s |
| Comparative | Muse Glimmer 30B | meta/muse-glimmer-30b via togetherai | 2026-10-05 01:53 | yes | 14 | 26 s |
| Budget constrained | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-05 00:16 | yes | 9 | 8 s |
| Budget constrained | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-04 22:56 | yes | 3 | 4 s |
| Budget constrained | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-05 01:38 | yes | 27 | 33 s |
| Budget constrained | Perplexity Sonar | sonar | 2026-10-05 00:00 | yes | 18 | 4 s |
| Budget constrained | Grok 4.1 Fast | spacexai/grok-4.1-fast-non-reasoning via vertex | 2026-10-05 00:19 | yes | 21 | 10 s |
| Budget constrained | Mistral Small | mistral/mistral-small via mistral | 2026-10-05 00:33 | yes | 5 | 4 s |
| Budget constrained | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-05 00:02 | yes | 25 | 54 s |
| Budget constrained | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-04 23:41 | yes | 5 | 3 s |
| Budget constrained | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-05 00:13 | yes | 19 | 26 s |
| Budget constrained | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-05 01:31 | yes | 15 | 40 s |
| Budget constrained | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-05 01:48 | yes | 25 | 259 s |
| Budget constrained | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-05 00:00 | yes | 25 | 60 s |
| Budget constrained | GPT-6 Luna | gpt-6-luna | 2026-10-04 23:18 | yes | 2 | 16 s |
| Budget constrained | Muse Glimmer 30B | meta/muse-glimmer-30b via togetherai | 2026-10-04 22:51 | yes | 18 | 29 s |
| Scale constrained | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-05 01:53 | yes | 9 | 9 s |
| Scale constrained | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-05 01:25 | no | 0 | 10 s |
| Scale constrained | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-05 00:42 | yes | 13 | 27 s |
| Scale constrained | Perplexity Sonar | sonar | 2026-10-04 22:43 | yes | 19 | 6 s |
| Scale constrained | Grok 4.1 Fast | spacexai/grok-4.1-fast-non-reasoning via vertex | 2026-10-05 01:29 | yes | 18 | 11 s |
| Scale constrained | Mistral Small | mistral/mistral-small via mistral | 2026-10-04 23:53 | no | 0 | 8 s |
| Scale constrained | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-05 01:11 | yes | 21 | 81 s |
| Scale constrained | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-04 23:33 | yes | 9 | 4 s |
| Scale constrained | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-05 00:24 | no | 0 | 30 s |
| Scale constrained | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-04 23:14 | yes | 24 | 32 s |
| Scale constrained | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-05 00:45 | yes | 13 | 22 s |
| Scale constrained | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-04 23:38 | yes | 5 | 19 s |
| Scale constrained | GPT-6 Luna | gpt-6-luna | 2026-10-05 01:30 | yes | 3 | 18 s |
| Scale constrained | Muse Glimmer 30B | meta/muse-glimmer-30b via togetherai | 2026-10-05 01:32 | yes | 13 | 21 s |
| Negative framing | Claude Haiku 4.5 | claude-haiku-4-5-20251001 | 2026-10-05 01:20 | yes | 17 | 10 s |
| Negative framing | GPT-5.4 mini | gpt-5.4-mini-2026-03-17 | 2026-10-04 22:37 | yes | 3 | 10 s |
| Negative framing | Gemini 3.5 Flash | gemini-3.5-flash | 2026-10-04 22:38 | yes | 20 | 25 s |
| Negative framing | Perplexity Sonar | sonar | 2026-10-05 00:30 | yes | 25 | 3 s |
| Negative framing | Grok 4.1 Fast | spacexai/grok-4.1-fast-non-reasoning via vertex | 2026-10-05 00:53 | yes | 24 | 13 s |
| Negative framing | Mistral Small | mistral/mistral-small via mistral | 2026-10-05 00:54 | yes | 5 | 4 s |
| Negative framing | DeepSeek V4 Flash | deepseek/deepseek-v4-flash via deepinfra | 2026-10-04 23:59 | yes | 23 | 52 s |
| Negative framing | Llama 4 Maverick | meta/llama-4-maverick via bedrock | 2026-10-04 23:03 | yes | 5 | 2 s |
| Negative framing | Qwen 3.7 Flash | alibaba/qwen3.7-flash via alibaba | 2026-10-04 23:48 | no | 0 | 37 s |
| Negative framing | Kimi K2 | moonshotai/kimi-k2 via novita | 2026-10-04 22:36 | yes | 25 | 27 s |
| Negative framing | GLM 4.7 FlashX | zai/glm-4.7-flashx via zai | 2026-10-04 23:10 | yes | 24 | 225 s |
| Negative framing | MiniMax M2.5 | minimax/minimax-m2.5 via minimax | 2026-10-05 00:41 | yes | 15 | 25 s |
| Negative framing | GPT-6 Luna | gpt-6-luna | 2026-10-05 01:03 | yes | 3 | 41 s |
| Negative framing | Muse Glimmer 30B | meta/muse-glimmer-30b via togetherai | 2026-10-05 01:47 | yes | 24 | 34 s |
Normalization in this category
Every judgment call made between the raw labels and the numbers above, listed so it is visible and reversible.
ShowHide
Category-scoped readings
Forcepoint read as Forcepoint Data Security Cloud
Lakera read as Lakera Guard
Lakera (by Check Point) read as Lakera Guard
Lakera (or CalypsoAI) read as Lakera Guard
Microsoft Defender read as Microsoft Defender XDR
Palo Alto read as Palo Alto Networks
SentinelOne read as SentinelOne Singularity
SentinelOne (Singularity Platform) read as SentinelOne Singularity
SentinelOne (powered by Prompt Security) read as SentinelOne Singularity
UnderDefense read as UnderDefense AI SOC
Unresolved, counted raw
AWS AI Security
Axiom Secure
CalypsoAI Moderator
Cato AI Security (formerly Aim Security)
Censinet
Check Point Software
Check Point Software (powered by Lakera)
Check Point ThreatCloud AI
Check Point — Lakera Guard
Cisco — Secure AI / Robust Intelligence
Cloudflare AI Security for Apps
CrowdStrike Falcon AI
CrowdStrike: Falcon AI security capabilities
F5 — CalypsoAI
Fortinet AI tools
Fortinet FortiAI
Google Cloud Chronicle (Security Operations)
Google Cloud: Security Command Center AI Protection
Google Security AI
HiddenLayer AISec Platform
Imperva AI Application Security
Lakera LLM Guard
Microsoft 365 Copilot with Purview
Microsoft Defender for Enterprise
Microsoft — Defender for AI / Microsoft Purview
Microsoft’s AI security stack
Microsoft’s AI security stack — Defender, Purview, Entra, Foundry, Agent 365
Netskope AI Security / AI Guardrails
Onyx Security
Palo Alto Networks Precision AI
Palo Alto Networks: Prisma AIRS / AI Gateway / AI Runtime Security
Premonition
Prompt Security — now part of SentinelOne Singularity
Securiti DataAI Command Center
SentinelOne Singularity Platform with Purple AI
Splunk Security Cloud
UnderDefense MDR
Varonis AI Shield
WitnessAI Unified AI Security and Governance Platform
Wiz AI Posture
Wiz AI-APP
Zscaler AI Access Security
Zscaler AI Security / AI Protect
Zscaler Secure AI Access (with AI Guard)
Discontinued, still offered
No shut-down product was recommended here.