SPLX (Zscaler) iconSPLX (Zscaler)

Automated AI red teaming, runtime guardrails and AI asset discovery for LLM apps and agents; formerly SplxAI, acquired by Zscaler in November 2025

1K+ GitHub Stars (Agentic Radar)
Zscaler Owner since Nov 2025
2023 Founded

Overview

SPLX (formerly SplxAI) builds security testing and protection for LLM apps and AI agents. Its products are Automated AI Red Teaming (simulated attacks such as prompt injection and jailbreaks against your apps), AI Runtime Protection (input and output guardrails), AI Asset Management (discovers models, AI workflows, MCP servers and guardrails and builds an AI bill of materials), AI Governance and Compliance, Dynamic Remediation that hardens system prompts from red-team findings, AI Model Security for stress-testing commercial and open-source LLMs, and the open-source Agentic Radar scanner for agent workflows. The company raised a $7M seed in March 2025 led by LAUNCHub Ventures, and Zscaler announced its acquisition on November 3, 2025. splx.ai still sells the SPLX-branded products with a "now part of Zscaler" banner.

The Verdict

Who Should Use SPLX?

Best For

  • Security teams that red-team chatbots and agents before release
  • Zscaler customers who want AI security from their existing vendor
  • Teams that need an inventory of models, agents and MCP servers
  • Companies that want to try red teaming on a free plan first

Not Ideal For

  • Buyers who need public prices for paid plans
  • Teams wary of product names and packaging changing during the Zscaler merge
  • Anyone wanting a fully open-source red-teaming tool

What's Great

  • Red teaming, runtime guardrails and remediation in one platform
  • Dynamic Remediation turns findings into hardened system prompts
  • Free starting plan plus SaaS, on-prem and VPC deployment options
  • Open-source Agentic Radar scans agent workflows
  • Backed by a large security vendor after the Zscaler deal

Watch Out For

  • Paid plans are quote-only
  • Integration into Zscaler is ongoing, so names and packaging may change
  • No public false-positive benchmark, per one buyer comparison
  • Very little independent user review coverage
  • Agentic Radar repo has not been pushed since November 2025

Pricing

View all features & details

Products

  • Automated AI Red Teaming
  • AI Runtime Protection (input and output guardrails)
  • AI Asset Management and AI bill of materials
  • AI Governance and Compliance
  • Dynamic Remediation
  • AI Model Security; Agentic Radar (open source)

Use Cases

  • Pre-release red teaming of chatbots and agents
  • Blocking prompt injection and data leaks in production
  • Finding shadow AI, agents and MCP servers
  • Evidence for AI compliance frameworks

How It Compares

Feature SPLX Mindgard Garak Promptfoo
Type Commercial platform (Zscaler) Commercial platform Open-source scanner Open-source plus cloud
Runtime guardrails Yes See vendor No See vendor
Asset discovery Yes See vendor No No
Free option Free plan See vendor Free Free

Sources: SPLX homepage · Help Net Security (Mar 2025) · Zscaler acquisition post

Is this your tool? Add a badge