Transparent pricing. No surprises.

Yes. €0 excl. VAT, no credit card required. On signup, 500 welcome credits are added to the first month's 100 monthly credits, then you receive 100 credits each month. Enough to run a first audit (5 scenarios × 3 AI models, the Discovery plan cap) on a real client.

One credit equals one query sent to an AI model. Per-audit consumption depends on the number of scenarios, the number of languages and the models you select, more advanced models consume more credits. Before each launch, the wizard shows an exact credit estimate; you can adjust scenarios or models before confirming. Each plan includes a monthly credit volume; unused credits do not roll over from one month to the next.

All displayed prices are in € excl. VAT. The VAT rule applied depends on your location and status, in line with the EU VAT Directive:

• Businesses (B2B) in the European Union with a valid intra-EU VAT number: reverse charge, no VAT invoiced; you declare it in your country.
• Businesses and private clients in France: French standard VAT applies (20%).
• Private clients (B2C) in another EU country: VAT of the customer's country, under the OSS (One-Stop Shop) scheme for cross-border electronic services.
• Clients outside the EU (United Kingdom, Switzerland, Norway, North America, rest of the world): VAT not applicable, service export.

Yes. Upgrades are immediate with prorated billing. Downgrades take effect at the end of the billing cycle. No commitment, cancel in one click from your dashboard.

Consultant (€79) suits a freelance consultant auditing their own clients. Consultant+ (€179) adds advanced features (showcase, scheduled audits). Agent (€349) opens the REST API and multi-account portal. Agency+ (€599) unlocks simultaneous multi-market audits, full agency branding and extra agents (+€199 per agent above 2). The table above details each line.

From Consultant+, your PDF reports carry your logo and brand colors. The client portal is a secure read-only link, no account required for your client. On Agency+, you personalize the subdomain, colors, and each agent on your team has their own workspace.

You can buy one-off credit packs (from 500 credits = €39 to 15,000 credits = €749 excl. VAT, approximately 2 to 68 minimal audits depending on the pack) without changing your plan. Beyond your balance, a technical overdraft of 100 credits is tolerated so as not to interrupt a running audit. Purchase is one click via Stripe, instant return to your space.

Yes. Infrastructure hosted in France, GDPR compliant. AI model queries route through our inference partners (full subprocessor list in our privacy policy). Your audits, analysis scenarios, reports and client data stay in our European database. Storage stays in Europe, evaluation requests run through our documented sub-processor (OpenRouter, GDPR art. 28), and no model training is performed on your data.

Your clients sell across several European markets, and each market has its own keywords, competitors and AI search habits. Instead of running 4 separate audits (one per language), the Agency+ plan runs a single audit that queries the models in each language with analysis scenarios adapted to the market, not a simple translation, but contextual localization. You get a cross-market dashboard and a Premium PDF with one section per language. Ideal for steering a European brand. See the GEO audit methodology.

The heatmap crosses your languages (FR / EN / ES / DE / IT / KO / JA) with the AI models queried and shows at a glance where your brand performs or underperforms. Each cell is colored by visibility score (green = strong presence, red = absent). You spot in seconds the language × model pair that needs the most attention, for example, if your brand is cited by ChatGPT in French but ignored by Gemini in German. Tied to the AI Share of Voice metric.

Yes. For every language that underperforms on the 6-dimension GEO scoring, an Anthropic Claude Haiku model analyzes your results and produces a prioritized action plan, specific to the market concerned, not a generic recipe. The scoring itself runs on AGS, our multi-judge engine that calibrates 5 AIs against each other before every audit. Recommendations cover content, Schema.org markup, E-E-A-T signals and citation strategy. You can export them as PDF or convert them into tasks in your project tool.

In native mode, the model answers with no external web retrieval enabled, an operational indicator of what the AI associates with your brand, not proof of presence in its training data. In web search mode, the model can query the web before answering, useful to compare internal knowledge vs fresh information. Full mode sends each scenario twice (once native, once web) to compare. Technical details in Native vs Web score.

AQA (AI Question Answer) is an open standard (MIT license) we created to make FAQs readable and usable by conversational AIs. Three compliance levels: Basic, Standard, Full. Adopted by the GEO community, it improves your chances of being cited by ChatGPT, Claude, Gemini or Perplexity when a user asks a question. More info on the AQA hub and in the complete guide.

Hero Grades summarize a brand's overall performance in a single letter, across the whole GEO scoring: presence in answers, quality of what is said and perception signals. A+ = dominant presence, cited and well described; F = nearly invisible. Each component has its own grade, and the overall grade is a weighted average. Detailed reading in the Hero Grades explainer.

Over 100 AI bot signatures (GPTBot, ClaudeBot, PerplexityBot, GeminiBot, etc.) are detected in real time on your client's site. You see which bots crawl which pages, how often, and which referrals come from conversational AIs (clicks from ChatGPT, Perplexity, etc.). Five open source collectors available (Go agent, WordPress plugin, PHP, WSGI, Cloudflare Worker), no personal data. Tutorial in the AI tracking guide.

A hallucinated URL is a URL invented by an AI but that doesn't actually exist on your site (e.g., ChatGPT recommends yourbrand.com/promo when that page doesn't exist). The result: a user clicks, hits a 404, and brand experience is degraded. The Brand Safety module automatically detects these hallucinations in responses, as well as associations with sensitive topics. Details in the glossary and in the full article on AI hallucinations.