Add native access for other models including FREE models and non-API free services eg Groq API, LLama3-70b , huggingchat, Poe
There is a program called ChatAll that gives access to many models (free, not free, chatgpt, openAI API, Claude API, Groq API, more).
https://github.com/sunner/ChatALL
Typingmind has a great UI - beats everything else, however the limited number of models is constraining it's use.
Could you please add native Groq API access (currently the API is free) so we can use llama3-70b (and others), and, if possible, other free options (eg. access to chatGPT(not the API)) - though free Llama3 access would be ideal.
ChatAll includes many models in their (sparse) interface that are accessed via the web (not API) including copilot, huggingface models, Poe. Adding these would be a game changer,
(I've found that llama3-70b is a very capable model, and for general queries is up there with claude3 and chatgpt4, so I'm using it for free (via huggingchat or groq.com) as a main AI and then using claude3/chatgpt4 via TypingMind for any queries that I need second opinions for.
Being able to use llama3 in TypingMind would be fantastic.


Log in to comment and vote
Comments10
Aug 7, 2024
Gold Bagel
Aug 7, 2024
Out of scope
Harlequin Meteorite
May 17, 2024
This would be amazing!
Of course, the interface of typingmind is FAR better than that of chatgpt (and everything else).
Since gpt-4o came out, I've been using the free version as my main AI, so not using typingmind much (though I'd prefer to use the TM interface).
If people could use our free/paid chatgpt accounts with typingmind, this would make typingmind FAR more popular.
And, it is possible, as the free github program ChatAll : https://github.com/sunner/ChatALL
provides this:
eg. https://raw.githubusercontent.com/sunner/ChatALL/main/screenshots/screenshot-1.png
Ngoc Nguyen
May 9, 2024
Hi, as of now, you can add Groq API and also other free options like DeepSeek, Ollama, LMStudio, DeepInfra as custom model on TypingMind. Does this work for your use case?
Sapphire Horizon
Apr 23, 2024
Wouldn't OpenRouter do that for you already, JT?
Harlequin Meteorite
Apr 28, 2024
Sure, open router is useful, however it is possible to access free models via hugging chat and other services would expand functionality.
Btw I have been using llama3 via groq and it is inferior to the output from hugging chat, which is a shame.
Maroon Telescope
Apr 28, 2024
Stephan Becker BTW I found a provider that gives access to llama3 cheaper than openai called DEEPINFRA, and it works with typingmind via an openAI compatible API
Ivory Alpaca
Sep 16, 2025
Vote for Poe, which has OAI compatible APIs now