{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreif2kg6cqun2zha725kbnzk43qincdgdawrw6zob7vmlmnap6n73ze",
    "uri": "at://did:plc:25rdn5elo5izoxrmtis34zuk/app.bsky.feed.post/3motxs64ox2j2"
  },
  "coverImage": {
    "$type": "blob",
    "ref": {
      "$link": "bafkreibgywtj2a3pxty3wmewga32k4j73ey5t56txdgsxdi3tb2uwy7wem"
    },
    "mimeType": "image/webp",
    "size": 419882
  },
  "path": "/vakeesh_moorthy_08edcca64/why-we-chose-agpl-instead-of-mit-for-neural-inverse-cloud-393l",
  "publishedAt": "2026-06-22T03:17:31.000Z",
  "site": "https://dev.to",
  "tags": [
    "ai",
    "cloudnative",
    "agpl",
    "neuralinverse",
    "https://cloud.neuralinverse.com"
  ],
  "textContent": "When we open sourced Neural Inverse Cloud, the easiest choice would have been MIT.\n\nMost developers like MIT. It's short, permissive, and widely adopted. If you've released an open-source project before, MIT is probably the first license you considered.\n\nWe didn't choose it.\n\nWe chose AGPL.\n\nNot because we dislike permissive open source. Not because we want to restrict users. We chose it because infrastructure software plays by different rules.\n\n##  The Infrastructure Problem\n\nMIT works incredibly well for libraries.\n\nYou publish code, developers use it, and occasionally improvements flow back into the project. Nobody is forced to contribute, but community norms often make it happen anyway.\n\nInfrastructure software is different.\n\nCloud IDEs, databases, developer platforms, deployment systems, and backend services can be monetized without ever distributing the source code.\n\nA company can:\n\n  * Fork your project\n  * Add proprietary features\n  * Launch a hosted version\n  * Build a competitive advantage on top of community work\n  * Never contribute anything back\n\n\n\nThe original project does all the R&D.\n\nThe fork captures the value.\n\nWe've seen this pattern repeatedly across open-source infrastructure over the last decade.\n\n##  Why AGPL Exists\n\nAGPL closes a loophole that traditional open-source licenses leave open.\n\nWith GPL, if you distribute modified software, you must publish your changes.\n\nBut what if you never distribute the software?\n\nWhat if you simply run it as a hosted service?\n\nThat's where AGPL comes in.\n\nIf you modify AGPL software and provide it to users over a network, you must also provide the source code for those modifications.\n\nThat applies to everyone.\n\nIncluding us.\n\nIf we improve Neural Inverse Cloud, those improvements stay open.\n\nIf someone else builds a SaaS business on top of it, their modifications stay open too.\n\n##  Why This Matters for Users\n\nWe wanted users to have guarantees.\n\nWith AGPL:\n\n  * You can self-host the latest version\n  * Community improvements remain accessible\n  * No company can create a permanently closed fork\n  * You always have an escape hatch\n\n\n\nThe software stays genuinely open.\n\nWith MIT, there's nothing stopping a company from taking the code tomorrow, adding proprietary features, and creating a version the community can never access.\n\nThat's not necessarily wrong.\n\nIt's simply not the ecosystem we wanted to build.\n\n##  The Enterprise Trade-Off\n\nLet's be honest.\n\nAGPL scares some enterprises.\n\nMany legal departments have blanket policies against copyleft licenses. Some procurement teams won't even evaluate AGPL software.\n\nWe're okay with that.\n\nNeural Inverse wasn't designed around enterprise procurement checklists.\n\nIt was designed for developers who want control over their tools and the freedom to self-host them.\n\nIf that means slower enterprise adoption, we're willing to make that trade.\n\n##  Competing With Ourselves\n\nOur business model is intentionally simple.\n\nThe source code is open.\n\nSelf-hosting is free.\n\nIf you don't want to manage infrastructure, we'll run it for you.\n\nWe charge for operations, reliability, infrastructure, scaling, and maintenance—not for access to the code itself.\n\nThat means we compete with our own self-hosted version.\n\nAnd we think that's healthy.\n\nOpen source should give users real choices.\n\n##  Why More Infrastructure Projects Should Consider AGPL\n\nAGPL isn't the right answer for every project.\n\nFor libraries, SDKs, and developer tools, MIT often makes perfect sense.\n\nBut for infrastructure software, AGPL creates something valuable:\n\nAlignment.\n\nThe incentives of the company, the community, and the users stay closer together.\n\nIf someone improves the platform, everyone benefits.\n\nThat's the kind of ecosystem we want to build around Neural Inverse Cloud.\n\nOpen source should be more than source code you can read.\n\nIt should be software that stays open—even when it's successful.\n\n**Neural Inverse Cloud**\n\n  * Self-hosted: github.com/NeuralInverse/cloud\n  * Managed Cloud: https://cloud.neuralinverse.com\n  * Free credit: $1.22 (no card required)\n\n\n\nWhat license would you choose for an open-source cloud platform: MIT, Apache 2.0, GPL, or AGPL? I'd love to hear the arguments from both sides.",
  "title": "Why We Chose AGPL Instead of MIT for Neural Inverse Cloud"
}