Developer Tools·Standings
LLM Security Tools
LLM security tools inspect prompts and model responses for risks; compare their defenses, deployment options, and pricing before choosing one.
Read the signal score out of 10 alongside the API, open-source, and free-tier marks. The board sets free-plan and entry-price details beside defenses for prompt injection, PII, secrets, output controls, model abuse, and deployment.
- 31 platforms
- 2 with a public API on this page
- 12 with a free tier
- Checked 23 Sept 2026
- 1 PIPrompt Inspectorpromptinspector.io
Prompt Inspector places a real-time security layer between untrusted input and an LLM application or agent.
- 2 NN
NVIDIA NeMo Guardrailsdeveloper.nvidia.comOpen-source Python package that adds programmable checks to block or validate LLM inputs and outputs.
- 3 CCyberSecEvalmeta-llama.github.io
Free tests for LLM cybersecurity behavior, including attack requests, code safety, and prompt attacks.
- 4 P
PromptGuardpromptguard.coSecurity layer that screens LLM requests for threats and masks sensitive data before processing.
- 5 SRSanctum Runtimesanctumruntime.com
Runtime trust layer that checks agent actions, supports approvals, and records policy decisions.
- 6 OGOpenlayer Guardrailsopenlayer.dev
Service that reviews AI traffic and can block, hide, flag, or send policy violations for review.
- 7 G
GuardionAIguardion.aiRuntime gateway that reviews agent and MCP actions, masks data, and stops unsafe tool calls.
- 8 VVotalAIvotal.ai
Runtime security platform enforcing AI guardrails, identity, data policies, and per-agent tool controls.
- 9 GA
Guardrails AIguardrailsai.comOpen-source Python framework for validating LLM outputs and mitigating PII, hallucinations, and jailbreaks.
- 10 IInferwallinferwall.com
Open-source firewall checking LLM requests and replies for attacks, leaks, and unsafe content.
- 11 GGLACISglacis.io
Runtime control service that constrains, pauses, or denies AI actions and signs decision records.
- 12 LGLLM Guardgithub.com
Python toolkit that checks prompts and replies for secrets, harmful language, leaks, and attacks.
- 13 WWitnessAIwitness.ai
Enterprise platform that maps AI apps and agent activity, applies policies, and records audit trails.
- 14 CCCisco Catalyst Centercisco.com
Cisco Catalyst Center is a centralized network controller and management dashboard for enterprise network operations.
- 15 VGVigil Guardvigilguard.ai
AI monitoring platform that checks prompts, responses, and agent actions, then allows, sanitizes, or blocks.
- 16 LA
Lasso AI Security Platformlasso.securityAI security platform discovering applications and agents, mapping risks, and inspecting prompts and tool calls.
- 17 RRocotalrocotal.com
AI governance platform that checks prompts and replies, applies guardrails, and sets user token limits.
- 18 PSPrompt Security AI Red Teamingprompt.security
AI security tool that tests homegrown apps for prompt attacks and data exposure, then adds runtime guardrails.
- 19 CA
Cloudflare Advanced Certificate Managercloudflare.comCloudflare Advanced Certificate Manager automates TLS certificate issuance, management, and renewal for websites.
- 20 GCGoogle Cloud Secret Managercloud.google.com
- 21 SGSecureAI Guardsecureaiguard.com
Server-side security layer inspecting model context for injection, data exposure, and risky agent actions.
- 22 HAHiddenLayer AI Guardrailshiddenlayer.com
AI guardrails platform for hosted or self-managed systems that blocks unsafe content, leaks, and prompt attacks.
- 23 CP
Check Point AI Guardrailscheckpoint.comAI Guardrails analyzes submitted LLM inputs and outputs to detect prompt injection, data leakage, policy bypass attempts, unsafe content, and other AI threats.
- 24 NS
Noma Securitynoma.securityPlatform discovering and governing agents, MCP servers, skills, models, and their data access.
- 25 ZAZeroTrusted AI Firewallzerotrusted.ai
Self-hosted AI firewall that inspects prompts, responses, and tools to block threats and control access.
Places #1, #2, #3, #4, #5 on this board are open to sponsors. Take #1 · Other ways to advertise
How to choose
Which deployment options fit?
Choose a deployment model that matches your setup: web, API, self-hosted, Linux, macOS, or Windows are the platform choices shown on the board.
If you need a browser extension as well as web and API access, Lasso AI Security Platform lists those options.
Which LLM risks does it cover?
Select controls according to the risks you need to address: prompt-injection defense, PII redaction, secrets detection, output guardrails, and model-abuse detection.
Review these fields together because a tool may address a different risk at each stage of a request or response.
Can I start without paying?
Check the free-plan, free-trial, and paid-from fields before committing. NVIDIA NeMo Guardrails, CyberSecEval, and Guardrails AI show no-cost plans, while PromptGuard combines a free plan with service starting at $19 per user per month.
GLACIS shows a free plan and a 14-day trial. VotalAI shows a free plan, pricing on request, and a 30-day trial.
Do I need open-source code?
When code access or modification matters, use the open-source indicator together with the pricing note. NVIDIA NeMo Guardrails, LLM Guard, Guardrails AI, and Inferwall mark their offerings as open source.
Does it connect through an API?
Use the API mark to identify tools designed for programmatic connections, then check the deployment model for how the security layer fits your applications.
Questions
Which threats can LLM security tools address?
The comparison covers defenses for prompt injection, PII, secrets, model outputs, and model abuse.
Which LLM security tools start with a free plan?
NVIDIA NeMo Guardrails, Prompt Inspector, CyberSecEval, PromptGuard, GLACIS, LLM Guard, Openlayer Guardrails, Guardrails AI, Inferwall, Sanctum Runtime, and VotalAI each show a free plan.
Which LLM security tools are open source?
Open-source code is listed for NVIDIA NeMo Guardrails, LLM Guard, Guardrails AI, and Inferwall.
How much does PromptGuard cost?
PromptGuard shows a free plan and a paid service beginning at $19 per user per month.
Which LLM security tools offer trials?
A 14-day trial appears for GLACIS, while VotalAI shows a 30-day trial.
Can I run LLM security tools on my own servers?
Self-hosted is one of the deployment options shown on the board. Check the deployment-model and platform fields for the tool you are considering.
Some top positions on this board can be paid placements, standing where they were sold. The facts on each platform are not for sale. How we rank · Disclosure