Incalmo

Incalmo builds reinforcement-learning cyber environments and security evaluations for AI labs and security teams to train and assess cybersecurity-capable models.

Sources:incalmo.ai lockstep.vc

incalmo.ai · updated Sep 29, 2026

RL environments Evals and benchmarks

Add to compareSuggest an edit

Headquarters: Menlo Park, United States
Source:sec.gov
Founded: 2025
Sources:cbinsights.com sec.gov
Headcount: 1-10
Source:linkedin.com
Total funding: $8.7M
Sources:sec.gov vcbacked.co
Latest round: Seed, $8.7M, June 2026
Sources:vcbacked.co sec.gov

Trust & data

How the vendor protects buyer material, where its data comes from, and where it can run.

Data sources: Synthetic / generated
Source:incalmo.ai
Quote
Synthetic / generated: Incalmo uses self-play and reinforcement learning to create extremely high-quality cyber environments to train models to be better at cybersecurity.
Not publicly listed: SOC 2, Trust center, Other certifications, Deployment, Exclusivity, Stated confidentiality practices

What they sell

Offerings: RL environments Evals and benchmarks
Sources:incalmo.ai lockstep.vc
Domains: Security
Source:incalmo.ai
Delivery formats: Docker containers
Source:github.com
Public samples & benchmarks: MHBench
Midnight Blizzard Network Red Team
Dumbbell Network Red Team
Star (PE) Network Red Team
Sources:incalmo.ai (1)incalmo.ai (2)incalmo.ai (3)arxiv.org
Not publicly listed: Named customers

Every change to these facts is kept in the public claims log. Record co_f3105aa9df9e.