Feature Requests

Could you add a feature to BoltAI that shows the changes the AI makes when it corrects spelling and grammar? It would be like a side-by-side or top-and-bottom comparison, similar to tools like Kaleidoscope or Beyond Compare. This way, users can easily see what was changed without having to copy the text into another program. It would save time and help people who don’t know how to use comparison tools. When configuring the prompt for inline mode, we would have an option to show the text diff in a pop-up window. From that window, we would have the option to edit the AI-modified version and then apply the changes, overwriting the original text. I quickly created a sketch of what the comparison window could look like. I don’t think any AI app has this feature yet, so it could make BoltAI stand out. What do you think?

Please add support for a RAG inside projects that we can query inside them. We use another tool for this and it's vital for our workflow to be able to query our knowledge bases without sending the entire file each time.

I really love the feature set and roadmap for this app. However, I'm really struggling with the UI and fluidity of the user experience. I'm primarily interested in the chat interface and working with files, and very interested in the developing 'projects' feature. I have been using another excellent chat app 'Chatwise' and though it is quite feature-rich, it lacks a lot of the file-upload and project capabilities that BoltAI has. However, Chatwise has the best chat-ui I've used. Clean, minimal and very fast. I know these things are subjective, but I wanted to include some screenshots to illustrate the display of information with both apps. As you can see, there is usually more info viewable on the screen, and (at least to me) is cleaner and easier to focus on. The biggest difference is obviously the more subtle bubble around user questions, and no bubble for assistant responses. It seems the padding is also a little less, but still feels more spacious. For someone like me (with ADHD) I find I'm able to process information and focus so much more with this UI. The base version of chatwise is free, so I recommend checking it out. I would love to move to Bolt, as it really feels like is has a great (and growing) feature-set, but I can't get past the state of the UI at this stage. Thanks! Greg.

It'd be awesome to be able to right click a conversation and re-run it in order, with all the same prompts, on a different model. This would be great to compare quality of output to evaluate models. It's cumbersome to do this by cutting and pasting every prompt in another conversation.

Please allow different sources to add files to a project. For instance, it would be great if you had an option for YouTube videos, an option for websites

Hi, Could you please add support for the new Elevenlabs conversational API that they launched yesterday? Thanks!

Now that Anthropic has enabled websearch via their API, it would be nice for Bolt to support it. I would like it with any other model that supports web search via their API.

One of the biggest selling points for Elephas, your competitor, to me, was the "Brains" feature. It allows you to add files, URLs, and even Notion documents and more as "background knowledge" to an Assistant. In my use case, I added some legal texts and statutes to a specific Assistant ("Brain" as they're called in Elephas), and whenever I chat with it, it remembers this background information (URLs must be cached, of course). It's more or less an advanced instruction system for Assistants. This feature would be a great addition, in my opinion.

Add the ability to customize prompts and models for generating summary titles.

Hi 👋 It would be good to have a button that allows you to synchronize all the files. Because it is sometimes long click on "Enable File Sync" for each file