Website reputation report

liquid.ai

Last checked 19 hours ago · Point-in-time assessment

● Evidence is limited

What is this website?

Public pages were observed on liquid.ai, but their subject was not confidently classified. An About page was also found. The sampled pages were marked as English-language. Recurring page-metadata topics include blog, device, on device.

No listed local patterns observedHomepage-only analysis is not a malware clearance. Read the scope and findings below.

76/100Safety Score
75%Evidence Confidence
Trust Radar evidence profile Eight evidence dimensions from zero to one hundred. Exact values surround the chart. Threat Intelligence55 Domain History88 Infrastructure95 Website Security95 Security Hygiene80 Transparency55 Reputation50 Operational Stability85
Trust RadarEight observed factors. Select one to inspect its evidence.

How to read this report

Safety Score evaluates observed signals. Evidence Confidence measures how complete, direct, fresh, and consistent those signals are.

No automated scan can guarantee that a website is safe.

Independent evidence brief

What we observed about liquid.ai

Updated Oct 1, 2026

The latest scan examined 10 public pages on liquid.ai. Website descriptions below come from the site itself. Registration, connection and threat findings were checked separately.

Purpose and visible content

On its About page, the site says: Liquid AI is building foundation models that respect privacy, performance, and domain, from architecture to inference.

Registrar and domain history

Registered on Dec 16, 2017. Registrar: Key-Systems GmbH.

Recorded expiry date: Feb 27, 2028. Registry data can change.

Connection and hosting evidence

The checked homepage returned HTTP 200 in about 1000 ms.

A valid TLS certificate was observed, expiring Nov 25, 2026.

Threat checks and score context

External malware-provider coverage was unavailable. The report does not treat this missing check as a clean result.

Safety Score: 76/100. Evidence Confidence: 75%. These measure different things.

Public content profile

Pages reviewed on liquid.ai

10 of up to 10 pages

We sampled public pages from this website to understand its subject and contact paths. The titles and short descriptions below come from the website itself. They are not endorsements or independent safety findings.

Liquid AI | Device-native foundation models.Liquid AI builds efficient Liquid Foundation Models (LFMs) for on-device, edge, and cloud AI with low latency, privacy, and hardware-aware deployment. Company | Liquid AILiquid AI is building foundation models that respect privacy, performance, and domain, from architecture to inference. Connect | Liquid AIGet in touch with the Liquid AI team. Tell us about your use case and we'll be in touch. From Cloud to Keyboard: LEAP Expands to Laptops with Cutting-Edge AMD Ryzen™ Power | Blog | Liquid AILiquid AI today announces native support for AMD’s latest Ryzen™ and Ryzen AI™ processors, unlocking new performance gains on-device AI solutions. LFM2.5-2.6B: Deploy Agents Everywhere | Blog | Liquid AILFM2.5-2.6B is an on-device agentic model that plans, calls tools, and runs multi-step tasks at 220 tok/s in under 2.5 GB. Open weights on Hugging Face. Reducing Doom Loops with Final Token Preference Optimization | Blog | Liquid AIAntidoom uses Final Token Preference Optimization to cut doom-loop rates in reasoning models by up to 90%, with minimal disruption to model behavior. Introducing LFM2.5: The Next Generation of On-Device AI | Blog | Liquid AIToday, we're excited to announce the LFM2.5-1.2B model family, our most capable release yet for edge AI deployment. It builds on the LFM2 device-optimized architecture and represen LFM2.5-Encoders: Fast at Long Context, Even on CPU | Blog | Liquid AIOpen-weight bidirectional encoders for classification, routing & NLU. LFM2.5-Encoder-230M & 350M deliver 8K context and fast CPU inference for edge and on-prem. Designing Loops for Production-Grade Work | Blog | Liquid AIWhat became standard practice after vibe-coding a production-grade BPE tokenizer trainer without reading a single line of code. LFM2.5-VL-3B: A Better and Faster Vision-Language Model for the Edge | Blog | Liquid AILFM2.5-VL-3B is Liquid AI's most capable vision-language model, with grounding, screen understanding, and function calling, fast enough to run on-device.

Public page evidence

Topics and keywords observed on liquid.ai

12 phrases

These phrases were extracted locally from sampled public page metadata. They are observations, not search-volume or ranking estimates.

blogTitle7 sampled pages
deviceTitle, H1, Meta description5 sampled pages
on deviceTitle, H1, Meta description5 sampled pages
edgeTitle, H1, Meta description5 sampled pages
modelTitle, H1, Meta description5 sampled pages
modelsTitle, H1, Meta description3 sampled pages
foundation modelsTitle, H1, Meta description2 sampled pages
loopsTitle, H12 sampled pages
fastTitle, H1, Meta description2 sampled pages
cloudTitle, H1, Meta description2 sampled pages
nativeTitle, H1, Meta description2 sampled pages
is anH1, Meta description2 sampled pages

At a glance

Key findings

✓

Report is current

The displayed evidence comes from the published immutable scan.

i

Confidence is separate

Unavailable evidence is shown as missing and never counted as clean.

!

Verify before sensitive actions

Check the exact URL and context before payments, credentials, or downloads.

Fully explainable

Why this score?

Versioned weights
Threat IntelligenceWeight 25%
55/100

Threat providers were unavailable. The neutral uncertainty baseline is not a clean verdict.

Domain HistoryWeight 10%
88/100

Domain registration age and expiry were checked.

InfrastructureWeight 10%
95/100

Observed DNS records and network address were checked.

Website SecurityWeight 15%
95/100

Observed certificate validity and TLS protocol were checked.

Security HygieneWeight 15%
80/100

Observed security header values were checked for basic protective settings. Empty values, disabled HSTS and unrecognized values receive no credit. This is not a complete browser-policy validation.

TransparencyWeight 5%
55/100

Observed site metadata and About page were checked. Publisher-supplied details are not independently verified.

ReputationWeight 5%
50/100

Popularity is weak context, not proof of safety. Unranked sites use a neutral baseline.

Operational StabilityWeight 15%
85/100

Observed HTTP status and final connection scheme were checked.

Server-side analysis

Automated security checks

Local antivirus and phishing checks

ClamAV · homepage HTMLLocal antivirus is not enabled
Phishing.Database · exact hostnameLocal feed snapshot unavailable

Sampled scripts and pages

Resource antivirus is not enabled

0 resources checked · 30 script links observed on the homepage. Partial resource coverage.

ClamAV inspects the fetched homepage HTML on our server. New scans also sample up to four linked scripts and two collected internal pages, with size and time limits. Scripts are not executed. Other downloads, private server files and logged-in pages remain outside this check. A signature match needs investigation and can be a false positive. No match is not a whole-site safety guarantee.

Exact hostname lookup uses a downloaded Phishing.Database snapshot. It does not send this domain to an external lookup API or mark sibling subdomains as harmful. Not listed does not mean safe. Snapshot age is download age, not proof that every entry was recently revalidated.

Homepage HTML only · Eight local rule families · local-security-v2

Checks use the fetched page and redirect chain. Scripts are not executed, forms are not submitted, and linked files are not downloaded. This is not a malware clearance.

Page resource inventory

Linked scripts30
Inline scripts67
Embedded frames0
External script hosts0

Counts describe the sampled HTML, not every resource loaded by the browser. External resources are common and are not automatically harmful.

Partial analysis. Only the first 512 KiB of the page were inspected.

No listed local patterns observed. The limited rule set cannot detect every threat or behavior.

External threat intelligence

Provider coverageNot checked

Local checks do not query external reputation databases. An unavailable check is never a clean result.

Registration

Domain information

Public RDAP observation
Canonical domainliquid.ai
RegistrarKey-Systems GmbH
RegisteredDec 16, 2017
Domain age8 years
ExpiresFeb 27, 2028
Registry statusesactive
Report statePublished

Infrastructure

Network & hosting

Observed endpoint, not legal location
IP address216.150.1.193
Autonomous systemUnavailable
Network organizationUnavailable
Observed countryUnavailable
CDNNot detected
Web application firewallNot detected

At a glance

Infrastructure summary

Observed during this scan

A compact view of the connection, DNS and security surface. These observations describe what the scanner could verify at scan time and do not prove ownership or ongoing availability.

HTTP response200 · https://www.liquid.ai/
Endpoint216.150.1.193
NetworkUnavailable
CDN and WAFNot detected · Not detected
DNS surface32 records · 2 nameservers · 1 mail exchangers
Protection signals5 security headers · 3 technologies · DNSSEC unknown

Page telemetry

Observed page statistics

Safe metadata only

These counters describe the fetched response and sampled HTML. Script counts describe references in the markup, not a complete browser network log. Cookie counts cover observed response headers.

Response size570,713 bytes
HTTP check duration1,000 ms
Cookies observed2
Script resources30 · 0 external hosts
Frames and sampled checks0 frames · 0 resources checked

Encrypted transport

TLS certificate

Valid when observed
StatusValid when observed
ProtocolTLSv1.3
IssuercountryName=US, organizationName=Let's Encrypt, commonName=YR2
Valid fromAug 27, 2026
Valid untilNov 25, 2026
Certificate namesliquid.ai

Resolution

DNS footprint

DNSKEY status unknown

DNSKEY presence does not verify the DNSSEC signature chain. Missing records can mean no record was returned or the lookup was unavailable.

A216.150.1.193TTL —
A216.150.16.193TTL —
NSaria.ns.cloudflare.comTTL —
NSzac.ns.cloudflare.comTTL —
MX1 smtp.google.comTTL —
CAA0 iodef "security@liquid.ai"TTL —
CAA0 issue "amazon.com"TTL —
CAA0 issue "comodoca.com"TTL —
CAA0 issue "digicert.com; cansignhttpexchanges=yes"TTL —
CAA0 issue "letsencrypt.org"TTL —
CAA0 issue "pki.goog; cansignhttpexchanges=yes"TTL —
CAA0 issue "ssl.com"TTL —
CAA0 issuewild "amazon.com"TTL —
CAA0 issuewild "comodoca.com"TTL —
CAA0 issuewild "digicert.com; cansignhttpexchanges=yes"TTL —
CAA0 issuewild "letsencrypt.org"TTL —
CAA0 issuewild "pki.goog; cansignhttpexchanges=yes"TTL —
CAA0 issuewild "ssl.com"TTL —
TXT1password-site-verification=JXRGODBLOBANJFWBORGXGXBNTQTTL —
TXTanthropic-domain-verification-r8jb12=C2FnuGKoVkS7B6WV284GepZtBTTL —
TXTapple-domain-verification=Syqo9PMBRLRIweECTTL —
TXTcentral-8695TTL —
TXTgoogle-site-verification=3foQk3llsfJj1dIrfSsLEeow93lhu4uGuLaKF7jvS4MTTL —
TXTgoogle-site-verification=B7JBDciW2myNY6a2keoA0KNGE06ytVb0zI2mzj04GGgTTL —
TXTgoogle-site-verification=oEt2IIaZpM96Ku3-uScd1HRW8tWs4Dy_YMJ9tU16s3MTTL —
TXTgoogle-site-verification=pJHTy9v6jKr6UGuwO6JCnLCEzk-yxPQMbmgbOZe4nj4TTL —
TXTinclude:44218040.spf05.hubspotemail.netTTL —
TXTneon-domain-verification=x-uxcv3CCpg96r1FDZ7Z0UeSa1gB7-fVTTL —
TXTopenai-domain-verification=dv-r3MBXgIoEb9mIg7sTwn87G82TTL —
TXTqk2gspncieTTL —
TXTv=spf1 ip4:18.7.71.111 ip4:18.7.71.112 ip4:18.7.71.113 ip4:18.9.1.111 ip4:18.9.1.112 ip4:18.9.1.113 include:_spf.google.com include:44218040.spf05.hubspotemail.net -allTTL —
TXTZOOM_verify_8BCy748r7J5p2Xc4WoKBKVTTL —

DNS observations · no external lookup API

Domain email protection

Exact-host SPF and DMARC observations help the owner investigate email impersonation risks. They are separate from website malware checks and do not change the Safety Score.

SPFPolicy record observed, not fully validated
DMARCPolicy record observed, not fully validated

What to do next

Review the records with your mail administrator. Inventory legitimate senders and test SPF and DKIM alignment before enforcing a stricter DMARC policy. Fix duplicate records and investigate overly permissive SPF rules.

This bounded check does not send email, probe SMTP, evaluate recursive SPF includes, verify DKIM signatures or apply organizational-domain DMARC inheritance. A missing subdomain record does not prove that no inherited protection exists. A DNS timeout is unknown, not an absent record. A policy record is not proof that messages authenticate successfully.

Responsible disclosure

security.txt

RFC 9116 observation

No readable security.txt file was observed at the standard locations. This is informational and is not evidence of malware.

Browser protections

Security headers

5 observed
permissions-policypresent camera=(), microphone=(), geolocation=()
referrer-policypresent strict-origin-when-cross-origin
strict-transport-securitypresent max-age=63072000; includeSubDomains; preload
x-content-type-optionspresent nosniff
x-frame-optionspresent DENY

Response headers

Cookie protections

Passive observation
Cookies observed2
Missing Secure1
Missing HttpOnly2
Missing SameSite0

Only cookie attribute counts are retained. Cookie names and values are never stored. These observations are not proof of compromise and do not independently change the Safety Score.

Detected stack

Technologies

Bootstrap 80% Next.js 96% Tailwind CSS 72%

Versions are shown only when publicly exposed by the page or response headers. An observed version is not, by itself, proof of a vulnerability.

JavaScript library advisories

JavaScript advisory analysis could not be completed or no eligible script files were sampled. No clean result is implied.

Recommended maintenance checks

Next.jsKeep Next.js and its runtime dependencies current, especially when server-side rendering is enabled.

About this website

What the site says about itself

Liquid AI is building foundation models that respect privacy, performance, and domain, from architecture to inference.

View the site's About page ↗

This is publisher supplied information. The independent safety findings are shown separately.

Publisher supplied

Public contact details

These details were observed on the website, not independently verified as an active support channel.

Structured facts

Website identity & response

Public page metadata
Page titleLiquid AI | Device-native foundation models.
First headingDevice-native foundation models.
DescriptionLiquid AI builds efficient Liquid Foundation Models (LFMs) for on-device, edge, and cloud AI with low latency, privacy, and hardware-aware deployment.
Canonical URLhttps://www.liquid.ai
Final URLhttps://www.liquid.ai/
HTTP response200 · text/html; charset=utf-8 · 1000 ms
Languageen
Structured typesOrganization · WebSite
Page page detectedhttps://www.liquid.ai/blog/lfm2-5-2-6bAbout page detectedhttps://www.liquid.ai/companyContact page detectedhttps://www.liquid.ai/connect?topic=enterpriseSupport page detectedhttps://www.liquid.ai/blog/liquids-edge-ai-platform-leap-expands-support-to-laptops-with-best-in-class-performance-on-amd-ryzen-tm-and-ryzen-ai-tm-processors
Local headers.permissions usb restricted100% evidence confidence
1100% evidence confidence
1100% evidence confidence
1100% evidence confidence
en100% evidence confidence
1100% evidence confidence
1100% evidence confidence
missing100% evidence confidence
Local headers.cors credentials with wildcard100% evidence confidence
Dns consistency.cname present100% evidence confidence
unavailable100% evidence confidence
1100% evidence confidence
2100% evidence confidence
32100% evidence confidence
2100% evidence confidence
Local cookies.cookies missing samesite100% evidence confidence
1100% evidence confidence
2100% evidence confidence
2100% evidence confidence
Local headers.cors present100% evidence confidence
Local headers.cors wildcard origin100% evidence confidence
Local headers.csp present100% evidence confidence
1100% evidence confidence
63072000100% evidence confidence
1100% evidence confidence
1100% evidence confidence
1100% evidence confidence
1100% evidence confidence
1100% evidence confidence
Local headers.permissions payment restricted100% evidence confidence

Connection path

Redirect chain

1 recorded hops
Hop 1https://www.liquid.ai/

The chain is shown exactly as observed by the scanner. A redirect is not itself a malware finding, but unexpected domains or an HTTP downgrade deserve review.

Premium monitoring

Catch downtime and security changes automatically

Regular checks, availability history, and alert emails are reserved for Premium websites. Free reports remain point-in-time scans and do not create recurring server work.

Enable monitoring for this website

Compared with the previous scan

What changed

A comparison will appear after the next completed scan.

Explainable matching

Digital fingerprint

fingerprint-v1
clear-radar-path-858

Comparable components help find technically related websites. A match does not prove common ownership.

DnsHashed private comparison value
TechnologiesHashed private comparison value

Material changes only

Change history

No meaningful changes recorded yet.

Website icon for liquid.ai

Visual evidence

Website screenshot

Screenshot of liquid.ai
Earlier browser capture retained because this scan did not produce a usable replacement. It is not fresh scan evidence.

For site owners

Show your verified trust report

Place this lightweight badge in your footer to show visitors an independent, dated WebsiteVirusScanner report. It can increase visitor confidence, but it never changes or guarantees the score.

Choose a badge style
Preview on your site
WebsiteVirusScanner trust badge preview

Use the current report link so visitors can verify the evidence. A badge is a transparency signal, not a paid ranking boost.

Questions about liquid.ai

FAQ

Still have questions? Contact us →
Is liquid.ai safe to visit?

The latest Safety Score is 76/100 with 75% Evidence Confidence. This is a point-in-time assessment of observed technical signals, not a guarantee. Check the exact address and avoid entering payment details if anything looks unusual.

Did the scan detect malware on liquid.ai?

External threat-provider coverage was unavailable for this scan, so malware status is unverified.

When was liquid.ai registered?

The public registry record shows Dec 16, 2017. Registration details may be privacy-protected or updated later.

Did liquid.ai have a valid TLS certificate?

The scanner observed a valid TLS certificate using TLSv1.3. A valid certificate alone does not establish that the site is safe.

How does liquid.ai describe itself?

The website's About page describes its purpose in the report above. This is publisher-supplied information and is separate from independent security findings.

Can a paid listing change this website's score?

No. Premium placement and owner-supplied pages are labelled separately. Technical evidence, the Safety Score and moderation decisions remain independent.