Suggest New Features 💡

think it's about time to upgrade the model. Since it's going to be redirected anyway, I think I just need to change the naming. https://docs.x.ai/developers/migration/may-15-retirement

Based on the rationale, Grok-4.3's hallmark of reasoning process disclosure was automatically applied, and the time mentioned in the official blog has already passed. We need to split Grok's model list into Grok-4.3-(low) and Grok-4.3-(high).

Win for everyone, as good as Pro (or better), but a lot faster and cheaper. https://deepmind.google/models/model-cards/gemini-3-5-flash/ Pro is coming bit later, but you could prepare for it already.

The current You.com chat interface has a user experience issue: pressing Enter once immediately sends the message. This creates a significant problem for users of languages that require character selection (such as Chinese, Japanese, etc.). Problem Description: Input methods for languages like Chinese require pressing Enter to select characters/words Currently, pressing Enter sends the message immediately, interrupting the character selection process Users cannot complete their input normally, severely affecting non-English user experience Suggested Solutions: Restore previous design: only send messages when the send button is clicked Implement a double-Enter mechanism: first Enter for character selection, two consecutive Enter presses to send Add settings option: allow users to choose Enter key behavior themselves This improvement would significantly enhance the experience for users of Chinese, Japanese, Korean, and other languages requiring character selection, making You.com more internationally friendly.

Implementing a language selection feature is crucial for optimizing user experience and reducing unnecessary costs. As it stands, both Claude models and the Genius mode default to English responses, even when users submit queries in their native languages. This often leads to users rephrasing their queries with explicit language instructions, resulting in additional token consumption and increased expenses.

I’d like You.com to offer an AI-powered documentation workspace, similar to Cursor for code, but focused on general documentation. I want to create and organize documents (e.g., PRDs, product strategy, company vision) in folders, reference them using ‘@’ mentions, and have AI generate and edit documents within this structure. Ideally, this would integrate with Visual Studio or offer a similar developer-friendly experience

It's their new flagship that replaces many other older models. Good time to update? https://mistral.ai/news/vibe-remote-agents-mistral-medium-3-5

Would be great to upload pictures and work with them. Also setting up quality and size, also 16:9 and 9:16. That information in a prompt didn't work.

The dialog is currently presented in a long text. It would be nice if the dialog could be displayed as a drop-down list, where the standard would be opened. The simplest order would be if one block was the user's question plus the answer from the LLM. This would make it much easier to navigate long discussions, as many answers from the LLM are so long that you no longer know what the question is. (Alternative idea: You always keep the current question visible as a sticky node. The problem here, however, is long questions). At the same time, you can think about having an outline area in a side window so that you can navigate over it.

The dialog is currently presented in a long text. It would be nice if the dialog could be displayed as a drop-down list, where the standard would be opened. The simplest order would be if one block was the user's question plus the answer from the LLM. This would make it much easier to navigate long discussions, as many answers from the LLM are so long that you no longer know what the question is. (Alternative idea: You always keep the current question visible as a sticky node. The problem here, however, is long questions). At the same time, you can think about having an outline area in a side window so that you can navigate over it.