LangWatch

license

Apache-2.0(core)+MIT(SDKs)

source

public

core-gated

gated(NOTICE and platform/app/ee/LICENSE.md put platform/app/ee under the LangWatch Enterprise License, covering SSO, SCIM, audit-log reading, governance, webhooks, billing, licensing, managed providers and back-office admin, each of which "verifies a license at runtime before it can be used")

An explicit open core: an Apache-2.0 platform, MIT SDKs, and an enterprise directory under a commercial license. The gating is read directly from the repository and does not rest on LangWatch selling a hosted tier. The NOTICE file names the split component by component - "platform/app/ee/ LangWatch Enterprise License (platform/app/ee/LICENSE.md)" against MIT for sdks/typescript, sdks/python and mcp/typescript, with everything else Apache-2.0. That license says the enterprise directory "implements the enterprise capabilities of LangWatch, including single sign-on, SCIM provisioning, audit logging, gateway webhooks, licensing, billing, and back-office tooling" and that "each verifies a license at runtime before it can be used". A runtime license check on shipped features answers this dimension directly. The enterprise license now also names governance, managed providers, event sourcing and webhooks alongside the SCIM, audit-log and billing modules.

• https://github.com/langwatch/langwatch recorded 2026-06-04

  Apache-2.0 core, MIT SDKs, langwatch/ee/ enterprise modules requiring commercial license

• https://raw.githubusercontent.com/langwatch/langwatch/main/NOTICE recorded 2026-08-12

  The license split, stated by the vendor per directory: "This product is licensed under the Apache License, Version 2.0 (see LICENSE.md), with the following components separately licensed in their respective directories: platform/app/ee/ LangWatch Enterprise License (platform/app/ee/LICENSE.md); sdks/typescript/ MIT; sdks/python/ MIT; mcp/typescript/ MIT." So the tier-setting licenses are Apache-2.0 and MIT, and the gate is one named directory.

• https://raw.githubusercontent.com/langwatch/langwatch/main/platform/app/ee/LICENSE.md recorded 2026-08-12

  The LangWatch Enterprise License. "The Enterprise Software implements the enterprise capabilities of LangWatch, including single sign-on, SCIM provisioning, audit logging, gateway webhooks, licensing, billing, and back-office tooling. The enterprise capabilities are the surfaces the software enables only for licensed deployments; each verifies a license at runtime before it can be used." It adds that some keep recording on an unlicensed deployment so no history is lost, audit logging in particular, and that "Reading, exporting, or otherwise making use of what they record is the enterprise capability, and that is what a license is required for."

• https://api.github.com/repos/langwatch/langwatch/contents/platform/app/ee recorded 2026-08-12

  What the enterprise license actually covers, listed rather than characterized: admin, audit-log, billing, event-sourcing, governance, licensing, managed-providers, saas, scim, sso, webhooks, plus LICENSE.md and README.md. The rest of the repository - cmd, pkg, platform, packages, sdks, mcp, charts, infra, services, plugins - carries no second license, so the published Apache-2.0 tree is otherwise the whole product.

No download or active-user figure is published for LangWatch, so the only honest signal is GitHub stars: 3,486 on the one declared repository, which bands at 1K-10K stars and level 2. Stars are a weaker instrument than downloads and cannot carry a product above level 3 on their own, and they have their own vocabulary rather than a download range, because a star is not a use. The count is read from the repository API and recorded with its digest; it drifts between reads, and drift on a star count is expected rather than a problem.

• https://github.com/langwatch/langwatch recorded 2026-06-04

  ~3.3k GitHub stars; no disclosed usage numbers

• https://api.github.com/repos/langwatch/langwatch recorded 2026-08-13

  stargazers_count = 3486 for langwatch/langwatch. No download, install, customer or user figure is published for this product anywhere the record cites, so stars remain the only signal.

Covers every dimension this category grades - OpenTelemetry-native tracing, offline evaluations and production monitoring, prompt management with GitHub integration, dataset creation and annotation, cost control through a built-in AI gateway with failover, and broad framework and provider integrations - plus distinctive agent simulation and scenario testing. That puts it level with Langfuse at 4, and it is held there because nothing here is a frontier-definer over the anchors.

• https://github.com/langwatch/langwatch recorded 2026-08-13

  README - "The platform for LLM evaluations and AI agent testing. We help teams test, simulate, evaluate, and monitor LLM-powered agents end-to-end", with headline sections for "End-to-end agent simulations" ("Run realistic scenarios against your full stack") and "Eval + observability + prompts in one loop".