EN
FR
ES
DE
What do AIs say about your clients?
Your clients are asking what AI says about them. That's normal: ChatGPT, Claude, Perplexity have become real intermediaries between brands and their future customers. We give you the tools to measure, understand and act — model by model, response by response.
No credit card, no commitment. Data hosted in Europe. 40% of internet users prefer asking an AI over a traditional search engine (Gartner 2025) — the shift is happening.
More than 50 models including ChatGPT, Claude, Perplexity, Gemini, Mistral, Copilot, DeepSeek, Grok and Llama — native and web search.
The Discovery plan is free with 600 credits included. Paid plans start at €79/month (Consultant).
Yes, white-label. We implement Schema.org, llms.txt, AQA, structured FAQ and IndexNow directly on your client's site, under your brand.
100% in Europe, GDPR-compliant. No storage outside the EU. AI queries are routed through public provider APIs (OpenAI, Anthropic, Google) under SCC art. 46 GDPR.
Agency or brand?
Choose your starting point
You're an agency, a consultant or a freelancer
Manage your clients' visibility in AI: multi-model audits, tracking over time and white-label reports.
Create my accountYou're a brand or a company
Discover for free how ChatGPT, Claude, Gemini and Perplexity talk about your brand, with no account.
Get my free pre-diagnosisOn AI Labs Radar, our service dedicated to brands.
Why now
The AEO Checklist to Rank in ChatGPT, Claude and Perplexity
Actionable tactics ranked by impact to make your site visible to generative AIs. Updated monthly with the latest patterns.
No spam. 1-click unsubscribe. GDPR-compliant.
For GEO agencies and consultants who report to their clients.
We show everything. About ourselves.
It's easy to sell audits. It's rarer to expose your own. Our dashboard, our client portal, our GEO showcase — all public, updated after every audit.
Our live dashboard
AGS scores, hero grades, mention rate, sentiment, position — unfiltered. Updated after every audit. The same dashboard, white-labeled, is deployable for each of your clients.
See the dashboardOur GEO showcase
Public page designed to be cited by AIs: long-form article, structured FAQ, Schema.org, hreflang in 4 languages. The same showcase, under their identity, is generated for each of your clients.
See the showcaseOur methodology
Hero Grades A+ to F, scoring across 6 dimensions, Wilson confidence intervals — every metric documented and explainable.
Read the methodologyDiscover AI Labs Audit in 3 minutes
Full platform walkthrough: launch an audit, interpret AI visibility scores and generate a professional report.
Six technical choices that change the game
Six engineering choices that change how trustworthy your score is — without hacks or magic promises.
Five AI judges that calibrate against each other
Each response is graded independently by five different models. The inter-judge reliability coefficient (GRC) is computed and published at every audit. You know exactly how defensible your score is — not just whether it looks good.
Score corrected for judge drift
Every LLM judge has its own bias — one is lenient (mean 3.4/5), another strict (mean 1.4/5). Our cumulative cross-audit EWMA calibration neutralizes that bias continuously, across the full audit history. When your score rises, it's your content — not the judge that happened to be drawn.
Cryptographic reproducibility
Every audit computes a cryptographic hash of its configuration: the pool of 5 LLM judges, scoring rubrics, AI models queried, prompts. If the hash changes between two audits of the same client, the dashboard displays a “Configuration changed” banner — you instantly know whether a score variation comes from your brand or from a methodology change.
Four European markets in one audit
With the Agency+ plan, a single audit queries models in French, English, Spanish and German simultaneously. Prompts adapted per market — not a simple translation. Cross-market heatmap to see where to act in one second.
Open AQA standard (MIT)
We created a standard to make FAQs readable by AIs. Open source, MIT-licensed, three compliance levels. AI tracking collectors are also open source — five available.
European sovereignty
Infrastructure hosted in France, GDPR-compliant (Regulation (EU) 2016/679). Your data stays in the European Union. No model training on your audits, no EU-outbound storage transfer.
GEO and AEO are disciplines less than two years old. No one can claim to know everything. What we know: conversational AI has become a real intermediary between brands and their future customers, and brands that don't appear are losing ground. What we do: we query each model in depth, analyze each response, and publish what we learn. Open standards, open source tracker, shared R&D. We're looking for a partner, rather than just a customer.
Standard AQA
An open standard we created to make FAQs readable by AI. 3 conformance levels (Basic, Standard, Full), MIT license.
MCP Server
Connect Claude, Cursor or Windsurf to the platform. 200+ tools to automate your audits, track your metrics and manage your clients from any AI assistant.
Open Source AI Tracking
100+ AI bots detected, 32 AI platforms tracked. 5 open source collectors: Go agent for Linux servers, WordPress plugin, PHP module, Python WSGI middleware, Cloudflare Worker. Secured with HMAC-SHA256. Zero personal data.
From advice to code
robots.txt, Schema.org, llms.txt, AQA: we support you through technical implementation. And if a feature is missing, we build it.
GEO Showcase Page
A dedicated web page per client, designed to be quoted by AIs: long-form article, structured FAQ, Google reviews, full Schema.org markup, 4-language hreflang, IndexNow. See it live on our own showcase.
Client Portal
Share a secure read-only link with your clients: full dashboard, AI visibility scores, history, action plan and bot tracking — all in your branding. Click to see it live.
Prompts that look like your market, not a dictionary.
A generic prompt measures generic visibility. To know what AIs really say about your clients, you have to ask them the questions their prospects actually ask. Our pipeline builds these prompts from each client's website and industry, classifies them by buyer-journey stage, rephrases them to stabilize the measurement, and validates them before every audit.
AI-assisted generation
From your client's website and industry, we generate a first set of relevant prompts. You adjust, add, remove — you remain the decision-maker.
Industry library
A base of 240+ prompts cataloged by industry and buyer intent (discovery, comparison, expertise, transaction), reusable and enrichable by you.
Classification by journey stage
Every prompt is automatically classified by buyer-journey stage — discovery, comparison, decision. You measure your coverage across the full funnel, not just the visible part.
Native multilingual
Your prompts translated and localized in FR, EN, ES, DE. One audit, four markets, side-by-side comparable.
Multiple reformulations
Every key prompt is asked under several different wordings. You get a distribution of mentions, not a single measurement — hence a statistically defensible score.
Pre-audit quality control
Every prompt undergoes a check on length, ambiguity and neutrality before being sent to the models. A poorly formulated prompt won't pollute your scoring.
A hyper-targeted sector prompt generates up to +78% more brand recommendations from LLMs (Princeton / Georgia Tech study, KDD 2024). That's exactly what our AI produces by drawing on your domain expertise — an asset you own and that can't be copied.
Open source contributor. European. Measurements that matter.
Not a tool we leave you alone with. Our team covers strategic GEO/AEO consulting and technical implementation. We intervene directly on your client's site, white-label, under your brand.
GEO consulting & strategy
Sector scoping, prompts generated by our AI from your domain expertise, competitive analysis, prioritized action plan per audit.
Technical implementation
Schema.org, robots.txt, llms.txt, AQA Standard, structured FAQ, IndexNow — direct execution by our team on your client's site, white-label.
Tracking & iterations
Scheduled audits (daily, weekly, monthly), full history and multi-period comparisons to track your AI visibility over time.
Every data point we produce is designed to be shared: PDF reports, client portal and public GEO showcase — all white-labeled under your identity. Every missing feature goes straight to our team: the roadmap is built with our users.
EU-hosted data · French team · GDPR-compliant
A partner, rather than a vendor
GEO and AEO are fields under construction. The rules change, models evolve, best practices are being defined as we go. We don't claim to have all the answers — but we're actively looking for them.
We publish our work: the AQA standard is open source, the AI tracker is open source, our analyses are documented. When we identify a pattern that works, we share it. When a client needs a feature that doesn't exist, we build it.
We're not looking for users who pay a license fee. We're looking for professionals who want to move forward on this field with us — test, measure, iterate, understand what works. If that's your approach, we're meant to work together.
What just changed in our platform
Four concrete advances that justify the price: multi-judge scoring, redesigned dashboard, continuous monitoring and simultaneous multilingual audit.
AGS v2 — Multi-judge scoring
Five LLM judges (Claude, GPT-4, Gemini, Llama, Mistral) vote on every answer. Wilson confidence, anti-drift, fully traceable. The score you can defend in front of a client.
Dashboard v4 — 5 structured zones
Diagnostic, AI Presence, Engine Performance, GEO Audit, Action Plan. One page, five zones, zero friction. Built to decide in under 5 minutes.
Automatic client website monitoring
Automatic daily cron across every site in your portfolio. If a site has been unreachable for more than 7 days, a KO badge appears on the client card with the date of the first failed check. Zero audit lost on a dead site.
Simultaneous multilingual audit (Agency+)
FR, EN, ES, DE in a single audit. Cross-market heatmap to compare AI visibility per language. One view to steer an international brand.
Frequently Asked Questions
How does an AI visibility audit work?
AI Labs Audit queries each AI model with your custom prompts and analyzes the responses. The system scores visibility across 6 dimensions, generates a Hero Grade from A+ to F, detects URL hallucinations, and produces a prioritized action plan. Results in minutes, white-label PDF report included.
What's the difference between GEO and traditional search?
Traditional search optimizes your visibility on Google. GEO optimizes your presence in generative AI responses. Both are complementary: good technical search also facilitates indexation by AI crawlers.
What is the AQA standard?
The AQA standard (AI Question Answer), created by AI Labs Audit, enriches your FAQs with freshness, source and maintenance metadata. It's the strongest signal to tell AI your content is reliable and up-to-date. 3 conformance levels: Basic, Standard, Full. Open source, MIT license.
What is GEO (Generative Engine Optimization)?
GEO is the discipline that measures and optimizes a brand's presence in conversational AI responses (ChatGPT, Claude, Gemini, Perplexity). AI Labs Audit queries 300+ AI models simultaneously and analyzes each response across 6 dimensions: citations, position, sentiment, coverage, source authority and semantic perception.
How much does AI Labs Audit cost?
5 plans from 0 to €599 excl. tax/month, depending on your client and audit volume. The Discovery plan is free (600 credits/month, welcome bonus included), no credit card required. Paid plans start at €79 excl. tax/month with white-label reports and scheduled audits. Transparent pricing, no commitment, change or cancel anytime.
Why choose AI Labs Audit over other tools?
AI Labs Audit is the only European platform combining AI visibility R&D, scientific measurement (not passive monitoring), open standards (AQA), 200+ MCP tools, complete REST API, and technical support. We actually query each model, we don't watch a dashboard. Data hosted in Europe, GDPR compliant.
Is traditional search dead with the arrival of GEO?
No. GEO is the natural extension of traditional search, not its replacement. Traditional search optimizes your presence on classic engines. GEO optimizes your presence in AI responses. A modern strategy covers both.