LibreChat
-
As a chatgpt user, I thought it would be cool to have the free equivalent that uses the Openai API. I don't know if this kind of project can be part of cloudron.
-
It actually is: https://docs.cloudron.io/apps/openwebui/
-
V Valexico referenced this topic
-
Looks like this has evolved a lot:
- https://alternativeto.net/software/librechat/about/
- https://www.librechat.ai/
- https://www.librechat.ai/docs/configuration/authentication
- https://www.librechat.ai/docs/configuration/authentication/OAuth2-OIDC
- https://www.librechat.ai/docs/features/agents
@girish What do you think?
Time to add some privacy-focused FOSS AI goodies to Cloudron?
-
What's the difference between opebwebui and librechat? Just wondering, I don't mind adding similar apps. But it seems they both can operate on an api key and are just frontends.
There's many apps in the pipeline already, so we have to wait a bit to get to this.
-
What's the difference between opebwebui and librechat?
LibreChat is supposedly more difficult to deploy than OpenWebUI. In which case, Cloudron would be able to add greater value by easing deployment of LibreChat.
Here is a quick summary via Grok:
"Here's a comparison between LibreChat and Open WebUI based on the available information:
Open Source:
LibreChat: Yes, LibreChat is open-source, available on GitHub. It uses an MIT license, allowing for free use, modification, and distribution.Open WebUI: Also open-source, with its code available on GitHub. It uses the Apache 2.0 license, which is similarly permissive.
Development Activity:
LibreChat: It shows active development, with regular updates and contributions from the community. It's noted for integrating with numerous AI providers quickly after their launch.Open WebUI: Has a healthy development pace with frequent commits and a focus on enhancing user experience and adding new features. The project's community seems engaged, with various pipelines and integrations being developed.
User Base:
LibreChat: While specific user numbers aren't detailed, it's popular among those looking for a customizable, enterprise-level AI chat interface, with a focus on privacy and control over data.Open WebUI: It's gaining traction for its ease of setup, user-friendly interface, and extensive features, particularly among those interested in self-hosting AI with control over their data.
Docker Support:
LibreChat: Offers robust Docker support for easy deployment, making it accessible for those familiar with containerization.
Open WebUI: Provides a Docker-based installation for both CPU and GPU environments, simplifying setup and deployment.Supported Features/Models:
LibreChat: Features include AI model switching, conversation branching, multimodality, and integration with various AI services like OpenAI, Azure, Anthropic, and Google. It supports RAG through integration with RAG API.Open WebUI: Known for its versatility, supporting models from Ollama and OpenAI-compatible APIs, with features like custom RAG, message monitoring, user rate limiting, and real-time translation. It also allows for the creation of custom models and real-time web browsing capabilities.
Retrieval Augmented Generation (RAG):
LibreChat: Integrates with RAG API for uploading files and asking questions, enhancing its capability for document-based Q&A. However, it lacks advanced file management features.Open WebUI: Offers an integrated RAG process that works well for small-scale document interaction, although it's noted to be less scalable for larger sets.
License:
LibreChat: MIT License.Open WebUI: Apache 2.0 License.
Pros/Cons:
LibreChat Pros:
Enterprise-level authentication and flexibility.
Supports a wide variety of AI models and services.
Strong focus on privacy with data retention control.
LibreChat Cons:
Configuration might be more complex for some users.
Lacks some of the administrative controls found in Open WebUI.
Open WebUI Pros:
Streamlined user management with a user-friendly interface.
Strong support for custom integrations and pipelines.
Excellent for those looking for an easy-to-set-up AI interface.
Open WebUI Cons:
Less granular user management compared to LibreChat.
RAG implementation might not scale well for large document sets.Both platforms serve different needs, with LibreChat leaning towards enterprise solutions and Open WebUI providing a more consumer-friendly setup with robust community support. Your choice might depend on your specific requirements, particularly regarding ease of use, control over AI models, and data privacy.
" -
@girish Open WebUI looks more focused on also hosting the open models, so that needs a GPU.
LibreChat is more for using APIs to externally hosted model providers.
I'd say LibreChat is the a more mature development, and appropriate for Cloudron to have first, as it will work with all existing hardware.
It's also one of those apps that pays for itself, in that the alternative of using Poe.com for a team of users is more costly than a Cloudron subscription and hosting.
-
@girish Oh, didn't notice that. Should come with a warning that you're about to download a massive image!
Yeah, LibreChat is might smaller and works without any local models.
I can see use-cases for both, but for most people I think they'll want LibreChat as a way to save on monthly costs for subscriptions to various AI models, as this enables just paying for API token usage, with no monthly fees. Once you see how it works, I think this would be the one you'd personally want to use as a daily driver for these things, too.
