Agenta

Partial info

Open-source LLMOps: prompt playground, evals, and observability.

Agenta is an open-source platform combining a prompt playground, versioning, evaluation, and observability, aimed at helping teams iterate on LLM apps collaboratively. It offers self-hosting and managed cloud.

License

Open source (MIT)

Deployment

Managed, Self-hosted

Pricing

Open-source self-host; cloud has free and paid tiers.

SDKs / languages

Python, Any (API)

Founded

2023

Strengths

Open-source, self-hostable
Playground + evals + observability
Collaboration-friendly

Limitations

Younger ecosystem
Breadth over depth in some areas

Alternatives

Braintrust

Eval-first platform for iterating on LLM quality.

Langfuse

Open-source LLM observability, evals, and prompt management.

LangSmith

Tracing, evals, and prompt management from the LangChain team.

Latitude

Open-source prompt engineering and evaluation platform.

Mirascope

Pythonic library for prompts as typed functions.

Portkey

AI gateway with observability, guardrails, and prompt management.

Compare all prompt engineering & management →

Facts verified 19 July 2026. Neutral summary, not an endorsement; verify current details with the vendor.