Removed all references to unfiltered or uncensored content from the app name, subtitle, keywords, and description Replaced screenshots, which contained references to unfiltered content, with new screenshots focused on the app's privacy features

- Improve accessibility - Allow selecting and copying any text

English (U.K.) - Updated models: GPT 5.5, GPT 5.5 Pro and DeepSeek V4 - Removed Anonymiser

- Anonymizer: Automatically replaces any personal information like phone numbers or addresses when you use 3rd party models like ChatGPT - Images support: Silo can see the images you upload and help you understand them (when using the default model, Kimi 2.5) - Updated to ChatGPT 5.4 - Plus users now have access to all models, same as Pro users - The keyboard slides down now when you send a message just like other chatbot apps

- Improved ZCash payments flow - Fixed Kimi 2.5 logo

- Upgrade the default model to Kimi K2.5 - Much smaller app bundle size - Fix unexpected retry issue if app was backgrounded during streaming

Bug fixes

- Extra usage via a streamlined (Qwen 3-30B) model when you hit your daily limit (Plus and Pro users) - iPad landscape support

- Extra usage via a streamlined (Qwen 3B) model when you hit your daily limit (Plus and Pro users) - iPad landscape support

Thank you for the feedback. We heard from many of you that the Plus rate limits felt too restrictive — so we increased them 5x for GLM 4.6. What's new: - 5x higher rate limits on GLM 4.6 - GLM 4.7 rolling out now (same higher limits) - Improved Problem Reports: better privacy controls + optional email if you'd like a reply Again, thanks for your feedback - keep it coming!

- Hotfix

- Import file feature now supports scanning text from images - New onboarding

- New model picker and privacy explainer page - Redesigned toolbar and navbar - Bug fixes

We’ve squashed some bugs and polished the app for a smoother, more reliable experience.

Thanks for all your feedback so far. In this release: • GLM 4.6 (TEE) is now the default model, a powerful open source model running in a Trusted Execution Environment for maximal privacy • Auto open new chat when you launch the app • We squashed a bunch of bugs!

- New model (GLM 4.6 - advanced reasoning and tool calling) - Deep Research (Pro feature) - Webapp with E2E-Encrypted sync (https://silo.freysa.ai) - Scheduled tasks (ask Silo to make one)

Enjoy faster messages, improved settings, and a brand-new onboarding flow. We’ve also fixed several bugs to keep things running smoothly.

New Pro plan now available. Starting with private on-device document analysis that keeps your sensitive files completely offline. More privacy-first Pro features coming next.

We’re now Silo! Our name has changed but our focus remains the same, building you a secure, private-by-design personal AI experience. What’s new in this version: - Faster web search: Faster and more comprehensive results when web search is used. - Enhanced chat experience: Improved performance in chats with faster search and smoother scrolling. - Edit & copy messages: Just long press any message to edit or copy it. - Scientific notation support: LaTeX rendering for complex equations and scientific notation. - Quality of life improvements: Added haptics and a new model selector.

- Small bug fixes

- Small bug fixes

- Small bug fixes

- Small bug fixes

- Small bug fixes.

- Small bug fixes.