Ask HN: Browser-UI with multi-LLM for the team?
I'm looking to provide our 4-person team access to various LLMs, including OpenAI, Anthropic, Meta's Llama, and some others.
I want to have browser-based UI (like ChatGPT) to upload files, get Markdown-reponses and have price control to avoid unexpected high costs at the end of the month. My current plan is to purchase API keys from different LLM providers and use LiteLLM as a proxy to setup cost limits OR to use OpenRouter.ai (potentially still using LiteLLM).
LibreChat currently doesn't support "passing through" uploads, but this feature is expected to be implemented this week: https://github.com/danny-avila/LibreChat/discussions/3760#discussioncomment-10445600
My Questions:
1. Is my approach using LiteLLM or OpenRouter with a UI a viable solution? I believe this could be more cost-effective than purchasing monthly chat accounts for each of the four users.
2. Are there other UIs similar to ChatGPT and Anthropic's interfaces, for the case that the new LibreChat attachment feature is too instable?
Update: It would be great to run a self-hosted open source stack. Are you looking for a shared experience like Google docs, or something like a document repo like a wiki? I don't use typingmind as a team, but they do have a team product and typingmind can be hooked into any LLM service. Is a good tool. Right now, I'm just looking for something that looks and acts like the ChatGPT-UI (lowest learning curve). Our team is very diverse and has no IT background. So we don't use collaborative tools (except MS Teams), but that doesn't stand in the way of productivity - it's powered by chaos. Thanks, I'll have a look at TypeMind. It's disappointing that the tool is closed source and that no modifications are possible or permitted. I briefly add in my requirements that I would like to have a self-hosted open source tool. I don't know about the team tool, but you can self host typingmind. I have been building alternative chat ui https://llmchat.co that might interest you. we doesn't support teams yet but we're building in that direction. btw it's open source and free to use. Thanks it sounds great. I'll definitely have a look at that. Maybe I can connect it to LiteLLM so that we can connect all kinds of models. Bionic - https://bionic-gpt.com Built around teams so you can build and share AI assistants based on your data. Thanks, I'll have a look at that. But anything with ‘book a call’ sounds unaffordable for us unfortunately ;) LibreChat is pretty good. Additionally, would love to chat about your use case. Seems really interesting. My co-founder and I are building out a solution to a similar problem. Ping me, email is in the bio. Let's discuss this publicly, please. Open Knowledge has the advantage that others also benefit from it. About the use case: Ultimately, I just want to enable my team to use different LLMs without having to create an account for each team member. However, it should be possible to upload attachments. RAG is not required at the moment. Thanks for the constructive push back.You are definitely right. How important is self-hosting in your private VPC vs a managed solution with good tenant isolation? How sensitive are these documents you plan to upload.