Since you're using the API, any particular reason why you didn't just use Librechat? It already has support for API keys for Claude, OpenAI, and you can also use it with locally hosted LLMs through Ollama, and has all of the same feature (saved conversations, file attachments, caching, etc.)
It uses the API so there are no "limits" to usage, it seems in this context that OP meant "limits" as in "I didn't like the limits of the UI" not the "usage limits". Also this will not work with a "normal" Claude subscription, it uses the API instead.
Ohhh... you're using the word platform to refer to only their web UI/website. I think that's confusing because when you say platform, I, and perhaps others, hear both the web UI and API, not just the web UI. Eg if someone is on the Stripe platform, they're using both the website and the API they provide. Since both of those things are run by Anthropic, I think of both of those as one platform coming from Anthropic (that they charge differently for).
Looks great. Curious if you created your own frontend interface mostly for it to be your own or because you found other options lacking? Librechat, OpenWebUI, etc.
I was just looking for exactly this today, as I don’t want to pay for Claude AI subscription and API separately and would like to run UI locally. Thank you!
Since you're using the API, any particular reason why you didn't just use Librechat? It already has support for API keys for Claude, OpenAI, and you can also use it with locally hosted LLMs through Ollama, and has all of the same feature (saved conversations, file attachments, caching, etc.)
https://github.com/danny-avila/LibreChat
Or the better-known Open-WebUI: https://github.com/open-webui/open-webui
Nice! I added this to the catalogue https://github.com/billmei/every-chatgpt-gui
How do you "bypass" their limitations? Please expand on how you were able to decipher their arbitrary limitations.
It uses the API so there are no "limits" to usage, it seems in this context that OP meant "limits" as in "I didn't like the limits of the UI" not the "usage limits". Also this will not work with a "normal" Claude subscription, it uses the API instead.
Since I'm using the API, I don't have to be limited to a specific number of messages.
But you're still subject to the limits set on their back end, no?
Well, the limitations for the API are different and much higher compared to using their platform
https://docs.anthropic.com/en/api/rate-limits
Ohhh... you're using the word platform to refer to only their web UI/website. I think that's confusing because when you say platform, I, and perhaps others, hear both the web UI and API, not just the web UI. Eg if someone is on the Stripe platform, they're using both the website and the API they provide. Since both of those things are run by Anthropic, I think of both of those as one platform coming from Anthropic (that they charge differently for).
Looks great. Curious if you created your own frontend interface mostly for it to be your own or because you found other options lacking? Librechat, OpenWebUI, etc.
I want to have more control over each chat I create, setting different temperatures, max output tokens, and personalities
I was just looking for exactly this today, as I don’t want to pay for Claude AI subscription and API separately and would like to run UI locally. Thank you!
You're welcome