in model selection drop down, mask models for which no api keys is provided (openai, gemini, etc.)
It takes too much spaces. Me for example, I only use openrouter. I have to scroooooollll wayyyyyy down to find my models, as model list are not (yet) : manageable/moovable/draggable group by provider or custom group (code/storrytelling/math/function call) filterable (by vision/function/context length/custom tag) Perhaps show just one per provider saying : "no api key" PS : a caveat : if accepted, where to annonce new models availability ? perhaps a small "new model available" ...
Saving tons of money on no rerunning prompts with "n" reponses
OpenAI can respond with multiple responses after sending one prompt. Example would be - I have a large large dialog, lots of tokens in use, and I am looking for a specific code solutions. But if gpt gives me poor code - I need to rerun it again and spend all that money on the prompt again. Or I can use something like n=3 and have 3 responses to try for the price of one prompt. Alternative use - if I need 10 ads written, Id rather use n=10 instead of asking for 10 ads in one response, because every next one is influenced by the previous one.
Multiple Stored API Keys with Aliases for easy selection / switching
Love the app so far! It would be fantastic to be able to have multiple API keys stored though (or to be able to point to some sort of secure store like an AWS Secret's Manager) where I could easily change out which API key I'm using. I use this for both work and personal so being able to make sure I'm using the right key and charging to the right OpenAI account (or any model account) would be greatly appreciated.
Chrome extension - for easy replies
Generate human like replies on linkedin, twitter, facebook, instagram, etc. Like https://magicreply.io/
OAI plugins, native code interpreter Model
Allow the use of OpenAI compatible plugins and the native code interpreter model
Custom Backup / Restore endpoint / Self Storage for Cloud Sync / Backup
Please add a customizable endpoint so we can do backup / restore automatically without using cloud.typingmind.com . This is so we don't have to export often manually and benefit from a more automated backup / restore feature without sending that data to the typing mind cloud. Endpoints would be provided as settings and it would be up to the user to implement their own solutions based on a simple documentation (eg: Backup, use Post, pass json + timestamp / Restore, use Get, etc)
Custom CSS in Settings
I've been using an Arc Boost with some custom CSS on TypingMind and have really been enjoying the option to add subtle styling to my personal preference. The new sidebar color selector got me thinking. Having a field under TypingMind settings where users could add custom CSS and sync it across devices would be a simple add that I think users would appreciate.
Typing Mind should use system theme
Under the Theme configuration option, there should be one called "System" to use system theme instead of manual switching. Probably good to hide the sidebar button when that option is selected.