Klira SDK

Developer tool

Open-source Python and TypeScript SDK that adds guardrails, tracing and evaluations to LLM applications and agents built on OpenAI, Anthropic, Gemini, LangChain, CrewAI and LlamaIndex, with policy enforcement, MCP protection and healthcare PHI anonymization; the SDK is Apache 2.0 licensed, and a cloud platform with Pro and Enterprise tiers is priced by contact.

Find alternatives to Klira SDK

TypeDeveloper tool

RoleStandalone tool

Bought forWatch an AI agent for suspicious or runaway behaviour

AvailabilitySold

DeploymentHybrid, Self-hosted

Intended forDeveloper, Enterprise

SecurityNot recorded

StatusActive

Sold withinSold on its own

PricingOpen source — official pricing ↗

Pricing, Sold within — getklira.com · 19 Oct 2026 Deployment — docs.getklira.com · 19 Oct 2026

What it does

Guardrails and safety
Evaluation and observability

Features 4

Guardrails, tracing and evalsOne SDK for guardrails, tracing and evaluations of LLM applications.
Adversarial probesRotating attack suites measure where an agent fails.
Tamper-evident monitoringSigned snapshots record what the agent was at every change.
PHI anonymizationRedact, mask, replace or hash PHI data in healthcare AI.

docs.getklira.com

Sources

Pricing

Official documentation · 19 Oct 2026

Terms: the SDK is Apache 2.0 licensed; the cloud platform has Pro and Enterprise tiers with no prices published, and the contact page is headed Demo & Pricing.

Description

Official documentation · 19 Oct 2026

Home page: "Guardrails, tracing, and evals for LLM apps."

What it does

Official documentation · 19 Oct 2026

Home page: "Guardrails, tracing, and evals for LLM apps."

Sold within