Giskard

Giskard belongs in the catalog because it covers a very practical and increasingly important layer of the AI stack: continuous testing and red teaming for LLM applications. The checked official site and pricing page position Giskard around open-source evaluation, continuous red teaming, LLM security testing, and RAG quality analysis. That makes it a strong fit for teams that want to catch hallucinations, prompt-security failures, and retrieval quality problems before and after deployment.

It also deserves inclusion because the packaging is honest and buyer-readable even though the commercial tier is enterprise-led. The checked official pricing page clearly distinguishes a Free open-source library from a quote-led Enterprise platform, then spells out enterprise capabilities like 50+ automated adversarial probes, RAG dataset generation, fine-grained quality metrics, CI/CD integration, and compliance-oriented controls. That is enough to write a premium catalog entry without making up self-serve pricing that the company does not publish.