Suggest New Features 💡

Now the clear official update to Mistral Large 2 is out, Mistral Large 3. https://mistral.ai/news/mistral-3

To the Security & Compliance Team, I am reporting a critical failure in your data retention logic that constitutes a severe privacy violation. Technical Issue: Despite the nonpermanent mode flag being explicitly enabled, all conversation data is persisting and remaining visible in the chat logs. The system is failing to honor the user's privacy directives. Compliance & Certification Concerns: This failure fundamentally undermines the validity of any claimed security certifications. If your architecture allows for data persistence when ephemeral processing is requested, your compliance posture is demonstrably flawed. This level of negligence calls into question the legitimacy of the auditing bodies that certified your platform. Formal Request & Escalation: Due to the severity of this data leak: DPO Contact: Provide the direct contact information for your Data Protection Officer immediately. GDPR Action: We are initiating a formal investigation regarding this breach of Terms of Service and potential GDPR non-compliance. Management Notification: This issue indicates a significant failure in your engineering and QA protocols. This correspondence serves as formal notice; senior management must be apprised of this systemic risk immediately. Reference Context: This appears to be a regression or escalation of a previously reported architectural flaw: https://yousearch.canny.io/bugs/p/nonpermanent-mode-personalization-status-loss I await your immediate response. For your information: Failure to take appropriate remedial action will result in further escalation. This includes public disclosure via LinkedIn and professional industry user groups, alongside the initiation of formal legal proceedings. Evidence & Consequences: Be advised that we have secured comprehensive forensic evidence of this behavior, including full network packet captures and proxied HTTP traffic logs. This documentation precludes any possibility of plausible deniability or obfuscation regarding this incident.

It would be desirable to have a single button to delete all chat history at once. Advatage: faster than deletibg each single chat

Advanced agents only available on mobile app, but agents require far more typing, which is inefficient on the mobile app

Create a URL for users to bookmark to always have private mode enabled such as private.you.com or you.com/private

Would be great to have a full privacy mode to use own files (PDF, txt, CSV, docx, xlsx) to work with. In Pro.

### Image-Based Search Functionality: Currently, You.com relies solely on text queries for search. Consequently, the capability to search using a combination of images and queries is lacking. As a result, when a user asks a question regarding an image, the image input fails to properly utilize the real-time access features. ### Previous Chat Search Functionality: The perceived accuracy is significantly low. I tested this by searching for a chat I had just yesterday, knowing for a fact that it existed. However, the search results either only retrieved very old chat records or brought up completely unrelated chat logs. ### Separation of Computing and Research Features: Currently, these features are integrated under the "Enable advanced research & reasoning" option in the Custom Agent settings. I suggest separating this into two distinct options: "Computing," which utilizes code execution and calculations, and "Research," which performs extensive searching. ### Relocating Web Search Settings to the Main Interface: Currently, the settings for SafeSearch, Region, and Time Period are located separately within the Web Search settings. I propose moving these settings to the main Web interface so that users can adjust them frequently while chatting. Furthermore, if these settings could be saved individually for each Custom Agent, it would allow for the creation of more personalized Custom Agents. [The image is an example] ### Sharing Custom Agents on the Webpage: Given the existence of a "custom-agents" thread on Discord, it appears that this is already somewhat planned. Once this feature is added, it goes without saying that users will be able to easily find and utilize agents created by other prompting experts. ### Replacing the Create Agent with Nano-Banana-Pro and Adding Image Editing Features: The currently used GPT-4o-Image-Generation model is outdated. I suggest replacing it with Nano-Banana-Pro, or at least Nano-Banana if cost is a concern. Additionally, it would be beneficial to have a feature that allows users to input images and edit them according to instructions.

Moving from Perplexity Spaces to You Agents! It's a very hard to organise and keep track of the searches done by the agent in the chat history.

have you.com respond in audio text to speech please

The recently released Nano banana pro version clearly surpasses the gpt-4o-image-generation model currently used by You.com in all aspects, including text generation, aesthetics, design, speed, and image editing performance.