Post

Has someone made an open source docker image that essentially deploys an open source ChatGPT interface and your favorite locally-hosted model to AWS or whatever? There are so many ChatGPT interface copycats out there, that there really should just be a common open source project that does this

27/1/25, 11:41 pm0 reposts1 like

nick

@nicktwoquick.bsky.social

28 Jan 2025

although this can't support local models afaik, this is close to what you are asking for i think partyrock.aws

Elliott

@techytacos.dev

28 Jan 2025

Not really what I think I'm looking for A self-hosted model is important because it should be easy to launch your own, private instance of "chatGPT" where the only middleman is the infra provider providing the VMs rather than these companies that just upcharge you on API calls to OpenAI / Deepseek

Elliott

@techytacos.dev

28 Jan 2025

Though I sorta doubt a personal, cloud-hosted instance of this is actually cheaper. Otherwise, maybe an open source FE that just gives you at-cost API calls is probably enough

nick

@nicktwoquick.bsky.social

28 Jan 2025

ah yes, seems sensible! im sure you have messed with this as a ui on ollama but in case not github.com/open-webui/o...

github.comGitHub - open-webui/open-webui: User-friendly AI Interface (Supports Ollama, OpenAI API, ...)User-friendly AI Interface (Supports Ollama, OpenAI API, ...) - open-webui/open-webui

Elliott

@techytacos.dev

27 Jan 2025

Hugging Face's GH org seems to have a lot of the parts, but not really an AIO setup w/ a FE