Fleuret AI (FLEURET AI SAS) · Fleuret: EU-Hosted AI Pentesting Agents for Web Apps and APIs
Fleuret is a Paris-based AI pentesting service whose agents map a web application you own or are authorised to test, its REST or GraphQL API and its external infrastructure, try to exploit what they find, and report only issues they can prove with a replayable, non-destructive proof of concept. It suits European SaaS, fintech and healthtech companies that need audit-ready pentest evidence for NIS2, DORA, ISO 27001 or SOC 2 at a flat €4,000 per web app.
Best for
European companies of roughly 50 to 1,000 people with a real web and API attack surface (SaaS, fintech, healthtech) and a NIS2, DORA, SOC 2 or ISO 27001 deadline, especially those that want pentest data and model inference kept on EU infrastructure
Not ideal for
Teams that need mobile pentests, or Active Directory or cloud pentests without first confirming coverage with Fleuret, organisations that require a self-hosted or open-source tool, financial entities looking for a DORA TLPT under TIBER-EU, and developers who want a free CLI to run security tests in CI themselves
Who it's for
CISOs, heads of security and CTOs at European SaaS, fintech and healthtech companies that need audit-ready pentest evidence
Fleuret is less a tool you operate than a priced pentest engagement delivered by agents: you hand over scope, test accounts and an authorisation, and get back proven findings in a ticketing workflow plus a report your auditor can file. That suits European mid-market teams whose real problem is compliance evidence on a deadline, and the EU hosting and open-weight models matter to regulated buyers. The open-source agents in this category give engineers more control and run in CI, but leave the reporting, scoping and re-testing to you. Before buying, get written confirmation of any scope beyond web apps and APIs, since Fleuret's pricing and FAQ currently disagree on Active Directory and cloud, size a yearly plan to the tests you will actually run because credits expire, and ask how much human review sits behind each engagement.
Who should use it
Security leads and CTOs at European SaaS, fintech and healthtech companies who need a credible, re-testable pentest of a web app and its API for NIS2, DORA, SOC 2 or ISO 27001, and who prefer a flat price and EU-hosted data over a multi-week consulting engagement.
Who should skip it
Teams needing mobile pentests now, or internal network, Active Directory or cloud pentests without confirmed coverage, organisations that must self-host security tooling, financial entities that need a TIBER-EU TLPT, and developers who want to run open-source security agents in their own CI.
Free
Free
Pentest
€4,000 (1 credit)
Billed one-time
Essential
€12,000/year (3 credits)
Billed annual
Continuous
€25,000/year (7 credits)
Billed annual
Posture
€50,000/year (15 credits)
Billed annual
Custom
On quote (20+ credits)
Billed annual
Free tier limits: The Free plan is self-serve, with no call or card, and covers 2 users and 10 repositories per organization with SAST, SCA, secret scanning, DAST and attack-surface discovery. It includes no pentest. Fleuret retired its paid Platform plan on October 28, 2026.
Note: Fleuret compares its €4,000 pentest with the €15,000–€30,000 it says a consulting firm typically quotes for the same surface. A web app, API, MCP server or AI system costs one credit, a multi-app or cloud scope two, and Active Directory two to three; re-tests of a finding are free. Extra credits are €4,000 on any plan, and yearly credits do not roll over. The modules listed per plan come from Fleuret's pricing; its FAQ still lists Active Directory and cloud pentests as not yet available. The free exposure check (Cassini) is passive and is not a pentest.
Available models
First SOC 2 or ISO 27001 audit at a SaaS company
A flat-priced pentest with a dated report and re-test gives the auditor vulnerability-testing evidence without scheduling a multi-week engagement.
NIS2 or DORA evidence for an EU-regulated company
The report is mapped to both frameworks, and findings, reports and model inference stay on Scaleway infrastructure in France.
Closing the loop on access-control bugs
A cross-tenant IDOR arrives as a ticket with the exact request and a replayable PoC, and the one-click re-test shows whether the fix holds.
Fleuret vs. Strix
Strix is an Apache-2.0 open-source pentesting tool you run yourself in Docker with your own LLM, built for CLI and CI use, with a paid cloud platform and self-hosted enterprise options on top. Fleuret is a hosted commercial service only: you supply scope, test accounts and a signed authorisation, and get a flat-priced pentest per web app with a DORA/NIS2-mapped report, re-test, and data kept on EU infrastructure.
Fleuret vs. Shannon
Shannon is an AGPL-3.0 white-box pentester that needs both the source code and a running copy of the app, and that you run against staging with your own model. Fleuret does not ask for source code: it tests the running web app, its API and external infrastructure from the scope and per-role test accounts you provide, as a managed service with validated findings, a ticketing workflow and a compliance-mapped report.
What is Fleuret?
Fleuret is a French AI pentesting platform. Its orchestrator, Émile, maps a web application and its API, dispatches specialist agents to attack it, and reports only findings it can confirm with a replayable, non-destructive proof of concept. Findings go into a workspace you can push to Jira, Linear or Slack, and the pentest ends with a report mapped to DORA and NIS2.
How much does Fleuret cost?
Since October 28, 2026 Fleuret sells pentest credits. The one-off Pentest is €4,000 for one credit: a grey-box test of one web app or API, business logic and multi-step chains included, with replayable proofs of concept, the signed report, and platform access with unlimited re-tests for 90 days. Yearly plans are Essential (€12,000 for 3 credits), Continuous (€25,000 for 7), Posture (€50,000 for 15) and Custom (on quote, from 20 credits at from €2,800 each). A web app, API, MCP server or AI system costs one credit, a multi-app or cloud scope two, and Active Directory two to three. Extra credits are €4,000, unused credits do not roll over, and a one-off pentest is credited in full against a yearly plan taken within 60 days. The free plan covers scanners for 2 users and 10 repositories and includes no pentest.
What can Fleuret test?
Fleuret's FAQ says it tests web applications, REST and GraphQL APIs, and external infrastructure today, and that Active Directory, mobile application and cloud pentests are not yet available. Its pricing, updated on October 28, 2026, nonetheless lists MCP servers and AI systems, Active Directory from the Continuous plan, and cloud and infrastructure from the Posture plan. No plan lists mobile testing. If you need anything beyond web apps and APIs, confirm coverage with Fleuret before buying.
Is it safe to run Fleuret against production?
Fleuret says its proofs of concept are read-only and non-destructive, demonstrating access without writing, deleting or degrading anything, and that every request is traced and shipped with the report. It also offers to test a staging environment instead. You must provide a signed authorisation covering the scope and testing window before any test.
Where is Fleuret's data hosted, and which AI models does it use?
Findings, reports and workspace data are hosted by Scaleway in Paris. Fleuret says it uses open-weight models (gpt-oss and Kimi K2.5) running on Scaleway GPUs in France, and that engagement data is not sent to a third-party model API such as OpenAI's.
Does a Fleuret pentest satisfy DORA or NIS2?
The report is mapped to DORA's resilience testing requirements (Articles 24 and 25) and to NIS2, and Fleuret describes it as evidence for ISO 27001 and SOC 2 as well. Fleuret says the report is audit evidence rather than a certification, and that it does not replace a DORA threat-led penetration test (TLPT) under TIBER-EU.
Is Fleuret open source or self-hostable?
No. Fleuret is a hosted commercial service, and no self-hosted deployment is documented. Teams that want an open-source pentesting agent they run themselves can look at Strix or Shannon.

Strix (OmniSecure, Inc.)
Engineering and AppSec teams that want open-source, AI-driven pentesting of their own apps and APIs in the CLI or CI, with an option to move to a managed platform
FreemiumKeygraph
Engineering teams with access to both the source code and a staging copy of their web app or API who want open-source, evidence-backed security testing in CI
Freemium
Hadrian Security B.V.
CISO-led security teams at mid-market and enterprise organizations (Hadrian names 300 to 10,000 employees) with a large, changing internet-facing estate across cloud and on-premises, who want validated external exposures instead of scanner noise and pentest evidence for NIS2, DORA, ISO 27001 or SOC 2
Paid
XBOW
Enterprise and mid-market security teams that own more web applications than they can pentest by hand and want exploit-validated findings, fix retesting and audit evidence across the whole portfolio, bought through sales or a cloud marketplace
CustomFleuret sits between a traditional pentest firm and an automated scanner. You hand it a scope (URLs, plus an API specification if you have one), test accounts for each role you want tested, and a signed authorisation that defines what is in and out of scope and the testing window; Fleuret says it never tests a system without written authorisation from its owner, and its terms limit the service to assets you own or are authorised to test. An orchestrator called Émile then maps the attack surface and dispatches short-lived specialist agents, each focused on one attack vector such as authentication, access control (IDOR), injection, business logic or privilege escalation, and retired after its mission. A separate validation engine turns a suspicion into a finding only after a controlled, read-only challenge confirms it is exploitable, and a Coverage Graph tracks which endpoints, parameters and auth contexts have been tested so the system can decide when the pentest is complete. Fleuret's launch announcement names a second agent, Champollion, alongside Émile, but the platform documentation only describes Émile's architecture. Fleuret describes the service as AI agents combined with human offensive expertise, and says a flat price works because most of the work is compute plus a bounded amount of expert time. Results land in a findings workspace rather than only a PDF: each finding has the exact request, the affected path, a severity, a replayable PoC and a remediation, can be pushed to Jira, Linear or Slack, and can be re-tested after a fix. Around the pentests sits Sibylle, the platform layer that maps an organisation's domains, repositories and cloud accounts and tracks each app's risk posture from its business profile, code activity, scanner and PR-review findings, and remediation progress, so it can recommend where and when the next pentest is worth running. Every plan, the one-off pentest included, ships a signed audit-ready report in PDF and JSON with traceability mapped to DORA, NIS2 and ISO 27001, and Fleuret's FAQ explains how it also serves as evidence for SOC 2, while stating that it is audit evidence, not a certification, and that it does not replace a DORA threat-led penetration test (TLPT). Data residency is a central part of the pitch: findings and reports are hosted by Scaleway in Paris, and the open-weight models (gpt-oss and Kimi K2.5) run on Scaleway GPUs in France, with no calls to third-party model APIs. Fleuret's funding announcement names Brevo, Stoïk and Yogosha among its customers. On October 28, 2026 Fleuret moved its pricing to pentest credits: a one-off Pentest is €4,000 for one credit, yearly plans run from €12,000 for three credits to €50,000 for fifteen, and a Custom plan starts at 20 credits, alongside a free self-serve scanner plan. A web app, API, MCP server or AI system costs one credit, a multi-app or cloud scope two, and Active Directory two to three. The tradeoffs: it is a hosted commercial service with no open-source or self-hosted option; Fleuret's own pages disagree on scope, since the pricing lists Active Directory, MCP server and cloud modules on the larger yearly plans while the FAQ, updated the same day, still says Active Directory, mobile and cloud pentests are not yet available, and no plan offers mobile testing; there is no public product documentation beyond the website and FAQ; and yearly credits do not roll over, so a plan should match the number of tests a team will actually run.
Fleuret replaced its Platform, Pentest one-off and Enterprise plans with pentest credits: a one-off Pentest at €4,000 for one credit, Essential at €12,000 a year for 3 credits, Continuous at €25,000 for 7, Posture at €50,000 for 15, and Custom on quote from 20 credits, plus a free self-serve scanner plan. Every plan now includes the signed audit-ready report and Sibylle, the platform's risk-posture layer. Enterprise and the paid Platform plan were retired.
Fleuret announced a €4M pre-seed round led by RAISE Ventures, with Auriga Cyber Ventures, Wind Capital, Better Angle and cybersecurity business angels. It says the money will fund hires in AI, software engineering and offensive security and a faster build-out of the platform, with the aim of moving pentesting from one-off audits to continuous testing.
Fleuret moved to three plans: a free-to-start Platform plan for scanners, a flat €4,000 one-off pentest per web app, and Enterprise on quote. The separate Advanced Pentest tier was retired on September 21, with its scope folded into the one-off pentest.
Are you the founder? Claim this listing →