NEW REPORT: You Enabled AI. Now What? The Nightfall Guide
Read more

Pick your Edition. Control How AI Moves Your Data

Schedule Your AI Exposure Assessment
Trusted by startups and Fortune 500 enterprises to control AI and how their data moves.
Chat with a Nightfall expert
Get a demo
Packages

Choose your edition

Foundation Edition

AI-Native Data security for humans and AI apps, across your whole organization.

Detect and stop sensitive data exposure across SaaS, AI apps, and endpoints. One detection engine, real-time blocking, deployed in about an hour.
$ per user/year
Includes
Endpoint agent, 2 devices per user
All supported SaaS apps, including posture and access revocation
AI-based classification and lineage
Shadow AI blocking on the endpoint
Nyx autonomous investigation
150 GB data-at-rest discovery included
Dedicated customer success manager
Priority support with 1-hour SLA
Best value - Most popular

Premier Edition

The full platform: AI-native data security for humans, AI apps, and AI agents.

Everything in Foundation, plus control over AI agents across IDEs, MCP servers, Claude Cowork and Claude Enterprise. Discover every agent, govern every tool call, block inside the session.
Choose your coverage:
Tier 1: $ user/year for up to 3 app
Tier 2: $ user/year for all supported apps
$ per user/year
Includes
All Foundation features
Hooks for Cursor, Claude Code (IDE and CLI) and VS Code on macOS and Windows. Scan and block prompts, MCP tool calls, tool responses and shell commands.
Local stdio and remote HTTP/SSE MCP discovery and inventory
Shadow MCP detection with per-server risk scoring
MCP Gateway policy enforcement
OpenTelemetry audit trail: cost, tokens, tool invocations
Claude Compliance API coverage for Claude Enterprise conversations, files, projects and activity feed. Same detectors as the rest of Nightfall: PII, PHI, PCI, secrets, source code, custom.
One policy set across endpoint, SaaS and agents, with a no-policy default. SIEM export to Splunk, Panther and Sumo Logic. Deploy via Jamf or Intune.

Model responses are logged, not blocked.

Optional add-on

Data add-on

Data Discovery and Classification Add-On

Perfect for businesses looking to remove inadvertent sensitive data exposure across SaaS apps.
Data pack tiers:
150 GB
Included
1 TB
$/year
3 TB
$/year
5 TB
$/year
20 TB
$/year

Compare Solutions

Feature
Foundation Edition
Premier Edition
Feature
Remove inadvertent sensitive data exposure across 12+ SaaS apps
Nightfall Complete
check - yes
Nightfall Complete
check - yes
Feature
Revoke data access misconfigurations
Nightfall Complete
check - yes
Nightfall Complete
check - yes
Feature
Email encryption
Nightfall Complete
check - yes
Nightfall Complete
check - yes
Feature
AI-based content classification
Nightfall Complete
check - yes
Nightfall Complete
check - yes
Feature
Prevent Shadow AI
Nightfall Complete
check - yes
Nightfall Complete
check - yes
Feature
Stop data exfiltration from Endpoints and Browsers 
Nightfall Complete
check - yes
2 devices per user
Nightfall Complete
check - yes
2 devices per user
Feature
AI-based data lineage
Nightfall Complete
check - yes
Nightfall Complete
check - yes
Feature
Nyx: AI-Powered Autonomous Investigation
Nightfall Complete
check - yes
Nightfall Complete
check - yes
Feature
Data discovery and classification (data at rest)
Nightfall Complete
check - yes
150GB included;
data packs for more
Nightfall Complete
check - yes
150GB included;
data packs for more
Hooks enforcement on Cursor, Claude Code, and VS Code (prompts, MCP tool calls, tool responses, shell)
check - no
check - yes
Local stdio MCP server discovery and inventory
check - no
check - yes
Remote HTTP and SSE MCP server discovery and inventory
check - no
check - yes
Shadow MCP detection with per-server risk scoring
check - no
check - yes
OpenTelemetry audit trail for Claude Code work
check - no
check - yes
Claude Compliance API monitoring for Claude Enterprise conversations, files, projects, and activity feed
check - no
check - yes
One policy set across endpoint, SaaS and agents, with a no-policy default
check - no
check - yes

Calculate your savings with AI-Native Data Security

See how much time and money you can save based on benchmarks from existing customers
Configuration
Adjust the parameters based on your current environment
Data violations per month
1000
0
5000
Manual investigation time
15 min
0 min
60 min
DLP analyst hourly cost
$100
$0
$500
Projected Savings
Your potential return on investment with AI-powered DLP
213 hours/month
Estimated Time Saved
$21,250
Monthly Cost Savings
6x ROI
Projected ROI
$255,000
Annual Cost Savings
* Calculations assume 85% reduction in manual investigation time through AI-based detection, investigation, and response.

Get started in
3 simple steps

Data Exfiltration Prevention

Connect in minutes

Connect your first SaaS app or deploy on endpoint
Data Exfiltration Prevention

Blocking from day one

Out-of-the-box policies and pre-trained ML detectors help prevent sensitive data exposure and exfiltration immediately
Data Exfiltration Prevention

Automated remediation

Automate response actions and notify end-users with self-remediation options

Frequently Asked Questions

How much does Nightfall cost?

Nightfall is priced per user, per year, in two editions. Foundation covers SaaS, endpoints and browsers. Premier adds AI agent security across IDEs, MCP servers and Claude. Volume discounts apply as user count grows, and both editions include 150 GB of data-at-rest discovery. Get a quote from sales, or start with a free 7-day proof of value.

What is the difference between Foundation Edition and Premier Edition?

Foundation is AI-native data security for people and the AI apps they use: API coverage for every supported SaaS app, an endpoint agent for exfiltration prevention, AI-based classification, lineage, and Nyx autonomous investigation. Premier is the full platform. It adds hooks for Cursor, Claude Code and VS Code, local and remote MCP discovery, shadow MCP detection, MCP Gateway policy enforcement, an OpenTelemetry audit trail, and Claude Compliance API coverage for Claude Enterprise.

Can I add AI agent security to Foundation for only some users?

Yes. If only your engineering org uses agents and MCP servers, you can buy Foundation for everyone and add AI & Agents for that group. Premier is the simpler option when most of the company needs agent coverage.

What is the difference between API-based SaaS coverage and endpoint coverage?

API coverage connects directly to a supported SaaS app and detects and remediates sensitive data inside it, including data already at rest. Endpoint coverage runs on the device and inspects data leaving it to any destination, whether or not that destination has an API. Both editions include both.

Does Nightfall offer a free trial?

Not a self-serve one. We run a free 7-day proof of value in your own environment, tuned to the policy that matters most to you, and you get a security assessment report at the end. If we find no meaningful data risk, the report is still yours.

How long does Nightfall take to deploy?

About an hour. Each SaaS app connects in minutes through its API, and the endpoint agent ships through Jamf, Intune or your existing MDM. Pre-trained detectors work out of the box, so there are no rules to write before you see results.

How many endpoint devices are included per user?

Two. Additional devices are available at the same per-user rate without changing editions.

What data types can Nightfall detect?

PII, PHI, PCI, secrets and credentials, source code, and anything you define with a custom detector. The same detectors run across SaaS, endpoints, email, and AI agent traffic.

How does Nightfall reach 95% precision?

Detection runs on machine learning models trained on labeled examples, not regex or keyword lists. The models read context, so a 16-digit number in a test file and a card number in a Slack DM are treated differently. Fewer false positives means fewer alerts your team has to clear.

Does Nightfall block data or just detect it?

Both. Policies can block, redact, quarantine, revoke access, or coach the user in real time, and you can scope each action by user group, data type and destination. Inside AI agent sessions, Premier blocks on prompts, tool calls, tool responses and shell commands. Model responses are logged, not blocked.

Does Nightfall stop Shadow AI and unapproved AI tools?

Yes. Foundation blocks sensitive data on the endpoint and in the browser before it reaches unsanctioned AI apps, and can redirect users to an approved alternative. Premier extends the same control to developer tools and agents: IDE assistants, MCP servers and Claude.

What is MCP security, and does Nightfall support it?

MCP (Model Context Protocol) is how AI agents connect to tools and data. It makes agent access standard, and risky. Premier discovers every local and remote MCP server, scores each one for risk, flags shadow servers, enforces policy on every tool call and at the gateway, and keeps an audit trail of what each agent did.

What does Premier add for AI agents?

Hooks for Cursor, Claude Code (IDE and CLI) and VS Code on macOS and Windows. Local stdio and remote HTTP/SSE MCP discovery and inventory. Shadow MCP detection with per-server risk scoring. MCP Gateway policy enforcement. An OpenTelemetry audit trail with cost, tokens and tool invocations. Claude Compliance API coverage for Claude Enterprise conversations, files, projects and the activity feed.

What does Nightfall integrate with?

SaaS apps including Slack, Google Workspace, Microsoft 365, GitHub, Jira, Notion, Salesforce and Zendesk. Identity through Okta and Google. MDM through Jamf and Intune. SIEM and alerting through Splunk, Panther, Sumo Logic and webhooks. See the integrations page for the full list.

What compliance certifications does Nightfall hold?

SOC 2 Type II and ISO 27001. Nightfall is HIPAA-ready and GDPR-compliant, and full documentation is available under NDA.

Who is Nightfall built for?

Security teams from startups to large enterprises, typically led by a CISO or director of security operations, at organizations from about 250 to 100,000+ users.

How does data discovery volume work?

Both editions include 150 GB of data-at-rest scanning to find and remediate sensitive data already sitting in your SaaS apps. To scan more, add a data pack: 1 TB, 3 TB, 5 TB or 20 TB per year.

Is there a minimum contract or user count?

Contracts are annual and priced per user, from small security teams to fleets of 100,000+. Talk to sales to confirm minimums for your size, or start with the proof of value.

Get a demo

Tell us a little about yourself and we'll connect you with a Nightfall expert who can share more about the product and answer any questions you have.
Not yet ready for a demo? Read our report:
‍The 2026 AI Agent Risk & Action Report