Open-source, cloud-native LLMOps platform for prompt management, observability, monitoring, caching and instant delivery of AI changes.
At a Glance
Self-hosted Pezzo LLMOps platform source code under the Apache 2.0 license.
Engagement
Available On
Listed Nov 2026
About Pezzo
Pezzo is an open-source LLMOps platform from pezzolabs, released under the Apache 2.0 license. It centralizes prompt design, version management, collaboration, observability and troubleshooting for AI operations. The latest GitHub release listed is v0.9.2, published 2024-05-15.
What It Is
Pezzo is a developer-first platform for managing and monitoring LLM-powered applications. Teams manage prompts in one place and deliver prompt changes instantly. They also observe and monitor AI operations to troubleshoot issues. The README says it can save up to 90% on costs and latency, which is a vendor claim.
Clients and Integrations
Pezzo provides a Node.js client (@pezzo/client on npm) and a Python client. LangChain is listed as supported. According to the README's client table, all three clients cover prompt management, observability and caching.
Self-Hosting Setup
Pezzo is built on open-source technologies: PostgreSQL, ClickHouse, Redis and Supertokens. The full stack can be run locally with Docker Compose. Development mode requires Node.js 18+ and Docker. It runs a server (NestJS/TypeScript) and a web console, with GraphQL codegen used during development. A hosted Pezzo Cloud option is also referenced.
Community Discussions
Be the first to start a conversation about Pezzo
Share your experience with Pezzo, ask questions, or help others learn from your insights.
Pricing
Open Source (Self-hosted)
Self-hosted Pezzo LLMOps platform source code under the Apache 2.0 license.
- Apache License 2.0 (OSI-approved)
- Self-hosted via Docker Compose
- Prompt management, observability, caching
- Node.js and Python clients
Capabilities
Key Features
- Prompt management and version control
- Instant delivery of prompt changes
- LLM observability and monitoring
- Troubleshooting of AI operations
- Request caching
- Team collaboration on prompts
- Node.js and Python clients
- LangChain support
- Docker Compose self-hosting
