Prompt & EvalOpen Source✦ Free Tier

DeepTeam

OSS red-teaming framework for LLMs and agents

⭐ 2,300 stars● Health 85/100 — Active· commit recency (40 pts) · star momentum (30 pts) · issue ratio (20 pts) · forks (10 pts)App Infrastructure

Open in Builder →Website ↗GitHub ↗

About

Adversarial red-teaming framework covering 50+ vulnerability types and 20+ attack methods, mapped to the OWASP LLM Top 10 and NIST AI RMF. Built by Confident AI, the team behind DeepEval, as a dedicated security-testing counterpart rather than a general eval library.

Choose DeepTeam when…

•You need to red-team an LLM or agent against a broad, named vulnerability taxonomy
•You want OWASP LLM Top 10 / NIST AI RMF mapped attack coverage out of the box
•You're already using DeepEval and want a matching adversarial-testing tool

Builder Slot

How do you know it's working?Optional for most stacks

Tests, evals, and experiment tracking to measure and improve your AI output quality

Dev Tools

Not applicable

App Infra

Recommended

Hybrid

Optional

Other tools in this slot:

PromptFoo DeepEval RAGAS Vellum PromptLayer Agenta TruLens Humanloop +1 more

Stack Genome Detection

AIchitect's Genome scanner detects DeepTeam in your project via these signals:

pip packages

deepteam

Integrates with (1)

DeepEvalPrompt & Eval

DeepTeam and DeepEval are built by the same team (Confident AI) and share infrastructure — DeepTeam supplies the adversarial attacks and vulnerability taxonomy, DeepEval supplies the metric/scoring layer.

→ Run adversarial red-team attacks through DeepTeam and score the results with DeepEval's metrics in one pipeline.

Compare →