Contact for Pricing

DeepKeep

AI-native security platform that firewalls, red-teams, and scans LLMs and computer-vision models against adversarial threats.

4.0 Security & Compliance AI
4.0 Very good 4.0
Covers both LLMs and computer-vision models, a wider scope than most LLM-only security tools No published pricing, so budgeting requires a sales conversation up front
Reviewed by Challenging Voice Editorial · Updated Jul 2026 How we rate
PricingContact for Pricing
Free planNo
CompanyDeepKeep
PlatformsAPI, Web
CategorySecurity & Compliance AI
Founded2021
Last reviewedJul 2026
Ask AI about DeepKeep ChatGPT Claude Perplexity

Overview

DeepKeep is an AI security company founded in 2021 by Rony Ohayon and a team of AI and cybersecurity specialists, headquartered in Tel Aviv, Israel. The company builds protection specifically for machine learning systems, an area traditional cybersecurity tools were not designed to cover.

The platform centers on an AI Firewall that inspects prompts and model outputs in real time, paired with automated AI Red Teaming that probes deployed models for jailbreaks and data leakage before attackers find them. DeepKeep extends this coverage to both large language models and computer-vision systems, and in 2024 raised $10 million in seed funding led by Awz Ventures to build out AI Usage Control and Model Scanning capabilities.

DeepKeep targets enterprise security and AI teams in regulated sectors such as finance, defense, and manufacturing that need guardrails around generative AI and agentic systems already in production. Pricing is not published; deployments are quoted directly through DeepKeep's sales team based on model volume and coverage scope.

Key features

  • AI Firewall that inspects prompts and outputs for injection and leakage attempts
  • Automated AI red teaming that probes models for jailbreaks before deployment
  • Model scanning for vulnerabilities across LLM and computer-vision systems
  • AI Usage Control for setting and enforcing organizational guardrails
  • Coverage extending to agentic AI systems, beyond single-model deployments
  • Recognition in Gartner and Omdia analyst coverage of the AI security category

Screenshots & demo

Demo video

Screenshots

DeepKeep screenshot 1

Pricing

DeepKeep uses custom pricing. Contact their team for a quote based on your needs.

  • Pricing modelContact for Pricing
  • Starting priceCustom
  • Free planNo
Visit DeepKeep

Pricing is provided as a guide. Check the official site for the latest plans.

Pros & cons

Pros

  • Covers both LLMs and computer-vision models, a wider scope than most LLM-only security tools
  • Automated red-teaming catches jailbreak paths before a model reaches production
  • Backed by seed funding earmarked specifically for platform expansion
  • Positioned for regulated industries where AI compliance documentation matters

Cons

  • No published pricing, so budgeting requires a sales conversation up front
  • Young platform (founded 2021) with a smaller public track record than incumbent security vendors
  • Enterprise focus means self-serve signup is not available
  • Effectiveness depends on integration depth, adding setup time before protection is live

How it compares

ToolRatingFreeFromBest known for
DeepKeep (this tool)4.0No—AI Firewall that inspects prompts and outputs for injection and leakage attempts
Adversa4.2No—Continuous automated red-teaming for GenAI apps and autonomous agents
Lakera4.3YesFreeReal-time prompt injection and jailbreak detection API
Dynamo AI4.3No—DynamoEval automated red-teaming across 20+ privacy and safety tests

Alternatives to DeepKeep

4 tools matched to DeepKeep on what they do, their category and their price.

Frequently asked questions

What is DeepKeep?
DeepKeep is an AI security company founded in 2021 by Rony Ohayon and a team of AI and cybersecurity specialists, headquartered in Tel Aviv, Israel.
Is DeepKeep free?
DeepKeep does not offer a free plan.
What are the best DeepKeep alternatives?
The closest matches in the directory are Adversa, Lakera, and Dynamo AI, compared side by side above.

Reviews

4.0 Editorial rating No reviews yet

Write a review

More tools like DeepKeep

Levo.ai

3.7

Runtime security for both traditional APIs and AI-native apps – discovers, red-teams, and firewalls your AI agents, LLMs, and MCP servers, not just your endpoints.

Contact for Pricing Security & Compliance AI