Comet ML

license

Apache-2.0(OSI)

source

public

core features withheld

yes — a self-hosted install lacks user management, which Comet's Enterprise plan carries with single sign-on, service accounts and view-only users

commercial

Comet-Enterprise(flexible deployments with single sign-on, service accounts and view-only users)

Opik's server, web application and evaluation components are Apache-2.0 and self-hostable, and the README calls the self-hosted build the full platform. Comet's self-hosting overview qualifies that: a self-hosted install has all Opik features but not user management. Comet's Enterprise plan carries it, with single sign-on, service accounts and view-only users. A commercial tier holding back part of the product makes this open core.

• https://github.com/comet-ml/opik recorded 2026-08-13

  Repository README, carrying the same statement - "Apache-2.0 licensed, free to self-host the full platform" - alongside the tracing, LLM-as-a-judge, dataset and integration sections. 21,353 stars.

• https://raw.githubusercontent.com/comet-ml/opik/main/LICENSE recorded 2026-10-08

  Repository LICENSE file, the full Apache License, Version 2.0 text with "Copyright (c) Comet ML, Inc".

• https://raw.githubusercontent.com/comet-ml/opik/main/README.md recorded 2026-10-08

  README states "Opik is licensed under Apache 2.0." and calls the platform "Apache-2.0 licensed, free to self-host the full platform".

• https://www.comet.com/docs/opik/self-host/overview recorded 2026-10-08

  Self-hosting overview: "you get access to all Opik features including tracing, evaluation, etc but without user management features".

• https://www.comet.com/site/pricing/ recorded 2026-10-08

  Comet pricing lists the open-source Opik build beside an Enterprise plan whose additions include "Service accounts and view-only users" and "Single sign-on".

Adoption is measured by PyPI downloads of the opik package, which ships more than 60 framework integrations for tracing and evaluation. The repository README discloses no trace-volume figure, so the reading is based on the download count alone, corroborated by GitHub stars on the repository.

• https://github.com/comet-ml/opik recorded 2026-08-13

  21,353 stars; README still describes extensive third-party integrations for observability

• https://pypistats.org/packages/opik recorded 2026-08-13

  Downloads last month 3,512,727 for opik (last week 770,341)

Opik covers OpenTelemetry-native tracing across many framework integrations, datasets and experiments with pytest-driven CI, LLM-as-judge metrics for hallucination, moderation, relevance and context precision, a prompt playground, and cost tracking. That matches the feature sets of Langfuse and MLflow, while Arize AX, a closed platform, goes further.

• https://github.com/comet-ml/opik recorded 2026-08-13

  README sections - "Track all LLM calls and traces with detailed context during development and in production", LLM-as-a-judge scoring, "Annotate traces and spans with feedback scores", datasets and "Extensive 3rd-party integrations for easy observability".

• https://pypi.org/project/opik/ recorded 2026-09-29

  PyPI page for opik: "any language with an OpenTelemetry SDK" can send traces to Opik, and evaluations integrate into CI/CD through PyTest.

• https://www.comet.com/docs/opik/ recorded 2026-09-29

  Docs home: "Deploy on your own infrastructure with Docker locally or Kubernetes at scale." and the navigation lists Cost tracking among the logging topics.