Giskard is an open-source testing and red-teaming framework for evaluating machine learning models, LLMs, and RAG pipelines against vulnerabilities like prompt injection and hallucination, founded in 2021 and headquartered in Paris, France. The Python SDK is free under the Apache 2.0 license, self-hostable, and runs automated vulnerability scans against an existing agent or RAG pipeline; it has accumulated roughly 5,700 GitHub stars. Giskard Hub, the enterprise platform, adds team collaboration, continuous red teaming, and scheduled testing on a quote basis. It targets ML and AI engineering teams that want an open-source-first path to model and LLM security testing.
ai-securityred-teamingopen-sourcellm-testing
Overview
- Founded
- 2021
- Pricing tier
- Low
- Startup-friendly
- Yes
- Enterprise-ready
- Yes
Agent security
- Prompt injection detection
- Yes
- Red teaming
- Yes
- Runtime protection
- No
- Model scanning
- No
- Open-source
- Yes
- Self-hostable
- Yes
- Free tier
- Yes: Open-source Python SDK (Apache 2.0) free and self-hostable; Giskard Hub enterprise platform is quote-based.