Vectra AI
Behavioural network, identity and cloud attack detection for security teams
Overview
Vectra AI is a network detection and response platform that uses machine learning to spot attacker behaviour across on-premises networks, identity, multi-cloud and SaaS, and ranks it into prioritised incidents. It added agentless cloud observability by buying Netography in 2025, launched Vectra AI Pro for AI-assisted security operations in August 2026, and sells through quotes and channel partners.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 19 Oct 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
Use cases
What you can produce with Vectra AI
- Behavioural detection of reconnaissance, credential abuse, lateral movement, command-and-control and exfiltration
- Prioritised incidents correlated across network, identity, cloud and SaaS
- Agentless cloud network observability for AWS, Azure, Google Cloud and Oracle Cloud from flow logs
- Exposure management with passive asset inventory, including unmanaged, IoT and OT devices
- Correlated signal for AI agents and analysts through Vectra AI Pro, plus an MCP server in early access
- Optional managed detection and response from Vectra
- A FedRAMP High authorised platform for US federal agencies
Vectra is one of the established names in network detection and response, a Leader in Gartner's 2026 Magic Quadrant and rated 4.8 from 472 ratings on Gartner Peer Insights. Its strength is behavioural detection that spans network, identity and cloud, ranked so a security team works a short queue, and the Netography purchase brought agentless flow-log coverage now reaching AWS, Azure, Google Cloud and Oracle Cloud.
The costs are commercial and organisational: no public pricing, a sales-led 45-day trial, and a product that pays off only with a security team able to act on it. Some managed-service customers report escalations that needed more first-pass analysis. A fit for mid-size and large enterprises; small businesses should look elsewhere.
What we found
Vectra AI is a San Jose network detection and response company, set up in 2011 by its own account, that models attacker behaviour rather than matching signatures. Its platform watches network traffic, identity systems such as Active Directory and Entra ID, cloud control planes and SaaS for reconnaissance, credential abuse, lateral movement, command-and-control and exfiltration, then ranks what it finds so a security team sees a short list of likely attacks instead of a long list of alerts. Vectra says it has more than 1,700 customers in 113 countries.
Gartner placed Vectra among the Leaders in its 2026 Magic Quadrant for network detection and response, for the second year running, and Vectra says it was positioned highest for Ability to Execute; the report itself sits behind Gartner's paywall. On Gartner Peer Insights the product averages 4.8 from 472 ratings, read on 19 October 2026. The review Gartner features as most favourable praises behavioural detection and reduced alert fatigue; the featured critical one, from December 2024, says the managed detection service escalated without enough first-pass analysis.
The cloud story has been assembled through acquisition. Vectra bought Netography in October 2025, bringing agentless observability built on cloud flow logs for AWS, Azure and Google Cloud; Oracle Cloud followed in June 2026. The January 2026 platform release added exposure management, and in August came Vectra AI Pro, which feeds correlated network, identity, cloud, SaaS, EDR and SASE signals to AI agents and analysts. An MCP server for AI assistants has been in early access since August 2025.
US federal buyers gained FedRAMP High authorisation in August and September 2026. In Singapore, Vectra partners with StarHub and appointed an Asia Pacific and Japan general manager in 2025. Nothing is priced publicly: buyers get a demo, a quote and possibly a 45-day trial deployment. It suits mid-size and large organisations with a working security team, not small businesses.
Compiled by RECATOOLS from public sources: vendor documentation and pricing, code repositories and published reporting.
Notable facts
- Vectra's AI models were trained on attack behaviour from over 10 million corporate networks, giving the models exposure to a wider range of attack patterns than any single organisation would encounter.
- The platform detects lateral movement — hackers moving between systems after the initial breach — with 97% accuracy according to MITRE Engenuity ATT&CK evaluations.
- Vectra was one of the first cybersecurity companies to apply deep learning to raw network packet data rather than relying on pre-extracted features.
Frequently asked questions
About this listing
This entry was compiled from publicly available data including Vectra AI's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Vectra AI unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to Vectra AI directly →
Spotted something out of date? Suggest an update →
More in Security & Safety
