{
  "id": 13104683,
  "title": "1 in 6 self-hosted AI apps ships a license GitHub can't classify",
  "url": "https://urgent.news/2026/10/09/1-in-6-self-hosted-ai-apps-ships-a-license-github-cant-classify",
  "topic": "ai",
  "section": "AI",
  "published": "2026-10-09T11:40:00.000Z",
  "source": {
    "name": "Dev.to",
    "slug": "dev-to",
    "url": "https://dev.to/archestack/1-in-6-self-hosted-ai-apps-ships-a-license-github-cant-classify-4ekm"
  },
  "original_language": "en",
  "account": "When considering building upon a self-hosted AI app, it is crucial to be aware that 16% of the 161 tracked applications ship a license GitHub cannot classify, representing around one in six. Typically, these apps adhere to standard licenses: Apache-2.0 on 63, and MIT on 57. However, 25 of these applications (the remaining 16%) utilize licenses that GitHub fails to categorize into any standard categories. This is often due to customized terms, dual licensing, or additional conditions appended to a normal license.\n\nProminent examples within this category include n8n, AutoGPT, Dify, and Open WebUI. It is important to review the license file before incorporating any of these applications into your own projects. While running these AI applications locally on your personal server typically does not raise concerns, distributing or selling access to them on a larger scale may necessitate a closer examination of the applicable terms.\n\nCopyleft licenses, such as AGPL-3.0 and GPL-3.0, account for a small fraction of the affected applications, with twelve and three programs respectively utilizing these licenses. Firecrawl, Khoj, and SillyTavern are among the applications employing AGPL-3.0, while the GPL-3.0 license is present in three other programs. The implications of these licenses only become relevant if modifications are made to the code and the altered version is offered to others over a network; for home servers, this scenario does not apply.\n\nConversely, 89 starter templates within the second category of self-hosted AI applications are permissive, with 70 following the MIT license, 17 adhering to Apache-2.0, and just two employing a BSD license. This alignment with GitHub's reported licenses makes it relatively straightforward to identify applications that warrant a closer examination: https://github.com/archestack/best-of-selfhosted-ai. The information presented here is compiled using assistance from AI tools, and the numerical data is derived from the public data associated with the list, as of October 9, 2026.",
  "summary": "Quick one if you're about to build something on top of a self-hosted AI app. We track 161 of them, and for most GitHub reads a standard license: Apache-2.0 on 63, MIT on 57. 25 of them (16%) ship a LICENSE file GitHub can't map to a standard license. Usually that means custom terms, a dual license, or a normal license with extra conditions added. Some big names are in that group: n8n, AutoGPT,…",
  "key_points": [
    "16% of 161 self-hosted AI apps have licenses GitHub can't classify.",
    "Common licenses are Apache-2.0 (63 apps) and MIT (57 apps).",
    "Notable apps include n8n, AutoGPT, Dify, and Open WebUI."
  ],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}