Dreadnode
Dreadnode builds AI-native infrastructure for security agents, offering the Dreadnode 2.0 platform plus Strikes, Spyglass, and Crucible products for AI red teaming, vulnerability assessment, and offensive security automation. It serves government agencies, AI labs, and enterprise security teams globally.
- Company typePrivate
- Founded2023
- HeadquartersBozeman, United States
- Headcount11–50
- GTM typeB2B
- OfferingSoftware
What Dreadnode does
Dreadnode is a Bozeman, Montana-based (Delaware-incorporated) software company founded in 2023 by Will Pearce and Nick Landers that builds AI infrastructure for the offensive security domain. The company operates a platform-plus-products architecture: the flagship Dreadnode 2.0 platform (launched March 2026) provides end-to-end agent lifecycle support (Build, Evaluate, Optimize, Deploy, Observe, Iterate) via API, function calling, and Model Context Protocol integrations, and is complemented by three commercial products — Strikes (cyber evaluation platform with Python SDK for testing AI capabilities and generating training data), Spyglass (AI red team toolkit for probing AI systems for vulnerabilities), and Crucible (AI hacking sandbox hosting CTF events). The underlying technology is model- and framework-agnostic and integrates with offensive security tooling such as Mythic/Apollo, BloodHound, and hashcat.
The company maintains a portfolio of open-source projects that drive community adoption and thought leadership, including Rigging (LLM interaction framework), AIRTBench (AI red teaming benchmark with 70+ challenges), Ares (multi-agent red/blue team system), and DreadGOAD (reproducible Active Directory lab). Revenue is generated via subscription to the Strikes, Spyglass, and Crucible platforms, with a freemium tier through Crucible and quote-based enterprise pricing for the other products. Go-to-market blends product-led growth (self-service SDK, free tier, GitHub distribution), enterprise field sales (government agencies, AI labs, large enterprise security teams), and community-led adoption (OAIC conference, Black Hat/Def Con presence). Customers span U.S. federal and defense agencies (accessed via an In-Q-Tel strategic partnership), AI labs and model developers, and enterprise security teams globally.
As of February 2025, the company employed approximately 15 people drawn from NVIDIA, Microsoft, Meta, Cohere, and NetSPI, and had raised $14 million in Series A funding led by Decibel with participation from In-Q-Tel, Next Frontier Capital, Sands Capital, Indie VC, Jon Oringer, and Aviso Ventures. Notable performance claims include an AI red teaming agent that achieved an 85% attack success rate across 68 adversarial goals in approximately three hours against Meta's Llama Scout model, and speedups exceeding 5,000x versus human operators on the most difficult AI security challenges.
Dreadnode firmographics
Firmographics- Name
- Dreadnode
- Legal name
- Dreadnode, Inc.
- Website
- https://dreadnode.io
- Company type
- Private
- Founded year
- 2023
- Operating status
- Operating
- Headcount range
- 11–50 employees
- Short description
- Dreadnode builds AI-native infrastructure for security agents, offering the Dreadnode 2.0 platform plus Strikes, Spyglass, and Crucible products for AI red teaming, vulnerability assessment, and offensive security automation. It serves government agencies, AI labs, and enterprise security teams globally.
- Ownership category
- akta.pro rank
Dreadnode industry classification
Industry- Product category
- AI Security Infrastructure
- NAICS
- Computer Systems Design and Related Services (5415), Other Computer Related Services (541519)
- SIC
- Services-Prepackaged Software (7372), Services-Computer Programming, Data Processing, Etc. (7370)
- akta.pro primary industry
- Model Security Testing & Red Teaming (adversarial ML, jailbreaks) (HDAAAKAC)
- akta.pro secondary industries
- Secure Model Deployment & Runtime Protection (sandboxing, isolation) (HDAAAKAH), App Security, Compliance & Review Automation Platforms (BPAMADAJ)
Keywords
Where Dreadnode is headquartered
LocationHeadquarters
- HQ city
- Bozeman
- HQ country
- United States
- HQ region
- North America
Offices1 record
Markets served
Dreadnode business model
Business model- GTM type
- B2B
- Offering type
- Software
- Cost components
- Personnel, Technology or R&D, Infrastructure, Marketing or Sales, Operations
Revenue model
- Strikes Platform Subscription: Build and execute cyber evaluations that both test AI capabilities and generate valuable training data for models and agents. Enterprise-grade platform with evaluation, observability, and agent development tools.
- Spyglass: AI red team toolkit that probes AI systems for vulnerabilities, enabling data-driven security decisions for safer model deployments. Provides simulated attack environments and vulnerability assessments.
- Crucible: AI hacking sandbox where practitioners test, learn, and advance their AI red team skills. Provides hosted environment for practicing AI red team skills with CTF challenges.
Pricing tiers
| Model | Billing | Price |
|---|---|---|
| Freemium | Monthly | Free tier available via Crucible platform |
Go-to-market motion3 records
Distribution channels4 records
Marketing channels8 records
Dreadnode product offering
Product offeringCore offering
Dreadnode provides an AI infrastructure platform that supports the full lifecycle of security agents — Build, Evaluate, Optimize, Deploy, Observe, Iterate. Its flagship commercial offerings are Strikes (a cyber evaluation platform with Python SDK for building and running AI security evaluations), Spyglass (an AI red team toolkit that probes AI systems for vulnerabilities), and Crucible (an AI hacking sandbox for practitioners). The platform is model- and framework-agnostic and is used by red teams, AI model developers, and government and enterprise security organizations.
Product overview
Dreadnode is an AI infrastructure company for the security stack, offering a platform-plus-modules architecture. The core platform (Dreadnode 2.0) supports the full security agent development lifecycle, complemented by three flagship products: Strikes (cyber evaluation platform with Python SDK), Spyglass (AI red team vulnerability toolkit), and Crucible (AI hacking sandbox). The platform is model and framework agnostic, supporting deployment via API, FN, and MCP protocols. Dreadnode also provides open-source tools including Rigging (LLM interaction framework), AIRTBench (red teaming benchmark), Ares (multi-agent red/blue team system), and DreadGOAD (AD lab environment). The platform enables security teams to build, evaluate, optimize, and deploy autonomous agents for offensive security tasks at machine speed and scale.
Differentiator
Problem solved
Functional benefit
Products and services
- Dreadnode Platform AI infrastructure platform that supports the full security agent development lifecycle — Build, Evaluate, Optimize, Deploy, Observe, Iterate. Model- and framework-agnostic, deployable via API, FN, and MCP. Built for security teams, AI model developers, and government organizations.
- Strikes Cyber evaluation platform for building and executing evaluations that test AI capabilities and generate valuable training data for models and agents. Includes a Python SDK (Strikes SDK), observability tools, visual analytics, and benchmarking for agent performance against security tasks. For security teams and AI model developers.
- Spyglass AI red team toolkit that probes AI systems for vulnerabilities and enables data-driven security decisions for safer model deployments. Provides simulated attack environments and vulnerability assessments. For AI model developers and enterprise security teams.
- Crucible
Quantifiable outcome
- AI agent achieved 85% attack success rate across 68 adversarial goals in 3 hours of testing against Meta's Llama Scout model
- +3 more outcomes
Companies that use Dreadnode
Customer profileNamed customers3 records
Segments4 records
Ideal customer profiles4 records
Dreadnode technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- Yes
- API docs
- API detail
Core technology
AI maturity
App detail
AI capability7 records
Feature10 records
Dreadnode partnerships and signals
Strategic signalPartnerships
Two partnerships are on record, tiered co-organizer.
- RemoteThreatco-organizerRemote Threat co-organized the inaugural Offensive AI Con (OAIC) conference with Dreadnode and DevSec. This is the world's first conference exclusively focused on the intersection of AI and offensive cybersecurity.
- DevSecco-organizerDevSec co-organized the inaugural Offensive AI Con (OAIC) conference with Dreadnode and RemoteThreat. Event took place October 5-8, 2025 at Seabird & Mission Pacific Resorts in Oceanside, California.
Scale indicators7 records
Recent moves6 records
Expansion highlights6 records
Dreadnode competitors and assessment
Company assessmentDirect peers
- Protect AI: AI/ML security platform covering model scanning, red teaming, and supply-chain security. Overlaps directly with Dreadnode's offensive AI testing and evaluation capabilities across the model lifecycle.
- SafeBreach: Breach and attack simulation platform running continuous adversary emulation. Closely parallels Dreadnode's platform concept of automated, agent-orchestrated offensive security testing at machine scale.
- AttackIQ: Breach and attack simulation and continuous security validation platform, aligned with Dreadnode's Strikes platform in automating adversary emulation and evaluating defenses at scale.
- Robust Intelligence: AI model security platform for testing, validating, and protecting AI applications from adversarial inputs. Directly comparable to Dreadnode's AI red teaming and vulnerability assessment capabilities.
- HiddenLayer: AI security company focused on protecting AI models and applications from adversarial attacks, evasion, and theft. Closely comparable to Dreadnode's Spyglass and Strikes offerings as a model security and AI red teaming platform.
- Pentera: Automated security validation platform that continuously emulates attacker techniques against enterprise environments. Highly comparable to Dreadnode's autonomous agent-driven penetration testing and vulnerability research.
Broad incumbents
- CrowdStrike: Global cybersecurity incumbent with Charlotte AI and offensive testing capabilities. Represents the large platform outcome Dreadnode could be acquired into or compete against as AI-driven security becomes mainstream.
- Microsoft AI Red Team: Microsoft's internal AI red team, originally built by Dreadnode's CEO. As Microsoft and other hyperscalers productize their internal offensive AI capabilities, they represent both a competitive threat and a potential ecosystem partner.
Emerging players
- Calypso AI: AI security and model firewall platform focused on testing and protecting generative AI applications. Adjacent overlap with Dreadnode in offensive AI probing, with complementary defense-side capabilities.
- Lakera AI: AI security startup focused on LLM guardrails, prompt injection defense, and red teaming tooling. Partial overlap with Dreadnode's adversarial testing and jailbreak research.
Market position
Strengths5 records
Weaknesses5 records
Competitive moat5 records
Key risks6 records
Key highlights6 records
Customer concentration
Dreadnode social profiles
Digital presenceDreadnode compliance and trust
Trust signalCompliance1 record
Dreadnode financial estimates
Financial estimateRevenue estimate
Valuation estimate
Dreadnode leadership team
Management profileNumber of profiles
Profiles4 records
Dreadnode funding detail
Funding detailFunding overview
Funding rounds1 record
Investors5 records
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
Dreadnode M&A and investment
M&A and investmentM&A
Investments
M&A and investment is available on the Subscription and Enterprise plan.Contact sales →
Frequently asked questions about Dreadnode
What does Dreadnode do?
Dreadnode provides an AI infrastructure platform that supports the full lifecycle of security agents — Build, Evaluate, Optimize, Deploy, Observe, Iterate. Its flagship commercial offerings are Strikes (a cyber evaluation platform with Python SDK for building and running AI security evaluations), Spyglass (an AI red team toolkit that probes AI systems for vulnerabilities), and Crucible (an AI hacking sandbox for practitioners). The platform is model- and framework-agnostic and is used by red teams, AI model developers, and government and enterprise security organizations.
Is Dreadnode a public or private company?
Dreadnode is a private company. It is classified as venture growth investor backed and is currently operating.
When was Dreadnode founded?
Dreadnode was founded in 2023. It employs 11 to 50 people.
Where is Dreadnode based?
Dreadnode is headquartered in Bozeman, United States, in the North America region.
How does Dreadnode make money?
Three revenue lines are on record. Strikes Platform Subscription is the primary driver. The others are spyglass and crucible.
Who are Dreadnode's main competitors?
Direct peers on record are Protect AI, SafeBreach, AttackIQ, Robust Intelligence, HiddenLayer and Pentera. Broad incumbents are CrowdStrike and Microsoft AI Red Team. Emerging players are Calypso AI and Lakera AI.
Does Dreadnode have an API?
Yes. Dreadnode provides a Python SDK called the Strikes SDK for building and running AI security evaluations. The SDK enables programmatic execution of cyber evaluations, agent harnesses, and AI red teaming operations. It supports logging metrics, parameters, inputs, and outputs for evaluation campaigns. The platform also mentions API and MCP (Model Context Protocol) as part of its capabilities for flexible and safe deployment. Developer documentation is at docs.dreadnode.io.
What industry is Dreadnode in?
Dreadnode's product category is AI Security Infrastructure. Its primary akta.pro industry code is HDAAAKAC, Model Security Testing & Red Teaming (adversarial ML, jailbreaks), with a secondary code of HDAAAKAH, Secure Model Deployment & Runtime Protection (sandboxing, isolation). Its NAICS code is 5415 and its SIC code is 7372.