Forwarded from Sophie NEWS π’
- Added a new command
/aiprovider to change the AI provider (see attached screenshot)- We now support Mistral AI β a European provider using open-weight models, offering top-tier data privacy!
- The default model is now ChatGPT GPT-5-mini
- AI Chat can now access Sophieβs documentation, in addition to Inbuilt Memory and Internet Searches β just ask her directly how to use specific features (see attached screenshot)
- Added a shortcut in /help to quickly start an AI chat
- Significant upgrades were made across the entire AI stack
- Added English (United Kingdom) π¬π§ language
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
π1
Forwarded from Sophie NEWS π’
Small follow-up [#Beta] update:
- Upgraded Mistral, the privacy-first European model, to Mistral Medium, now set as the default model
- Mistral Small is now used for AI Translator when the selected AI Provider is Auto or Mistral
- Markdown parser now supports de-parsing Markdown headings (Mistral really loves those)
- AI Translator now displays translated text inside a collapsible blockquote
- Updated Chinese (Simplified) with community-submitted fixes
π As always, this update is rolling out first to users with Beta mode enabled and will gradually reach more users in the coming weeks through an A/B release process, ensuring the seamless experience.
- Upgraded Mistral, the privacy-first European model, to Mistral Medium, now set as the default model
- Mistral Small is now used for AI Translator when the selected AI Provider is Auto or Mistral
- Markdown parser now supports de-parsing Markdown headings (Mistral really loves those)
- AI Translator now displays translated text inside a collapsible blockquote
- Updated Chinese (Simplified) with community-submitted fixes
Please open Telegram to view this post
VIEW IN TELEGRAM
Forwarded from Sophie NEWS π’
Hey folks,
π₯ We're starting with a brand new Sophie wiki.
In short, we have an ongoing project at OrangeFox Recovery aimed at replacing the wiki engine, developed on our own, called FoxWiki.
You can read more about FoxWiki on its documentation website: https://foxwiki.orangefox.tech
Today we're starting to use it for Sophie as well.
β‘οΈ Check it out: https://sophie-wiki.orangefox.tech/
Continuing, there are many internal changes to the bot as well, from upgraded libraries to monitoring improvements.
A few bugs were fixed as well.
β¨ Lastly, regarding the AI changes, we've tweaked the AI moderator and AI translator to yield fewer false positives.
π As always, this update is rolling out first to users with Beta mode enabled and will gradually reach more users in the coming weeks through an A/B release process, ensuring the seamless experience.
The brand new wiki is available for all users.
In short, we have an ongoing project at OrangeFox Recovery aimed at replacing the wiki engine, developed on our own, called FoxWiki.
You can read more about FoxWiki on its documentation website: https://foxwiki.orangefox.tech
Today we're starting to use it for Sophie as well.
Continuing, there are many internal changes to the bot as well, from upgraded libraries to monitoring improvements.
A few bugs were fixed as well.
The brand new wiki is available for all users.
Please open Telegram to view this post
VIEW IN TELEGRAM
Today: Improved custom theming for foxwiki and a countless fixes for responsiveness (mobile UI).
Before and after.
Before and after.
If you're using my Bitwarden instance, this message might be relevant to you:
[Action required] bw.owo.family becomes bw.orangefox.tech
Hi name,
As the registered user of the bw.owo.family Bitwarden instance, we're letting you know that we're currently migrating to OrangeFox's infrastructure; this means the domain address will change. The new domain is bw.orangefox.tech
You're advised to log out and log in using the new domain name by 23.10.2026. Until then, you can keep using both domain names, as they will both be fully functional.
You are NOT required to do any data migration.
The Bitwarden instance remains open for new registrations as well.
We will repeat this email a few times in the following year.
For any questions you can mail admin@orangefox.tech
- Yacha
β€2
Today, I've released a first tagged version of the foxwiki, many improvements included (read on the screenshot) :D
R1 is already deployed on the https://foxwiki.orangefox.tech, the Sophie's wiki will be updated with the next release of Sophie.
I am looking forward to migrate the OrangeFox's wiki to foxwiki as well.
R1 is already deployed on the https://foxwiki.orangefox.tech, the Sophie's wiki will be updated with the next release of Sophie.
I am looking forward to migrate the OrangeFox's wiki to foxwiki as well.
β€3
Hey, everyone! Today I finished FoxPaste.
The idea is a paste website that keeps the content in focus.
The editor is based on FoxWiki's code elements, so it will automatically recognize the language and colorize it.
The editor is also smart. It often behaves like an IDE, showing hints while you type and properly supports changing intention with the Tab key.
Try it already - paste.orangefox.tech
The idea is a paste website that keeps the content in focus.
The editor is based on FoxWiki's code elements, so it will automatically recognize the language and colorize it.
The editor is also smart. It often behaves like an IDE, showing hints while you type and properly supports changing intention with the Tab key.
Try it already - paste.orangefox.tech
π₯°2
The first FoxPaste update!
Read the changelog on FoxPaste itself:
https://paste.orangefox.tech/68ded0a53d9cbe550c5671af
Read the changelog on FoxPaste itself:
https://paste.orangefox.tech/68ded0a53d9cbe550c5671af
β€2
Forwarded from Sophie NEWS π’
We are pleased to announce a significant update to our AI request routing system.
Starting with [#Beta], all AI requests will be centralized through OpenRouter, replacing the previous method of routing messages to individual AI providers. This change enables us to manage AI providers more efficiently and streamline the overall process.
OpenRouter will serve as the primary platform for routing AI requests, improving operational efficiency.
However, requests to the Mistral provider will continue to be routed through Mistral AI servers, ensuring that all data remains within the EU.
We are also excited to introduce Anthropic, featuring the Claude 4.5 model, which will be available for use in our /ai chat features.
Use /aiprovider to change your AI Provider.
Our Privacy Policy has been updated to include OpenRouter and provide detailed information on how and when data is routed to each provider or service.
π As always, this update is rolling out first to users with Beta mode enabled and will gradually reach more users in the coming weeks through an A/B release process, ensuring the seamless experience.
Starting with [#Beta], all AI requests will be centralized through OpenRouter, replacing the previous method of routing messages to individual AI providers. This change enables us to manage AI providers more efficiently and streamline the overall process.
OpenRouter will serve as the primary platform for routing AI requests, improving operational efficiency.
However, requests to the Mistral provider will continue to be routed through Mistral AI servers, ensuring that all data remains within the EU.
We are also excited to introduce Anthropic, featuring the Claude 4.5 model, which will be available for use in our /ai chat features.
Use /aiprovider to change your AI Provider.
Our Privacy Policy has been updated to include OpenRouter and provide detailed information on how and when data is routed to each provider or service.
Please open Telegram to view this post
VIEW IN TELEGRAM
β€3
Hey, people - I've challenged myself to build a UI mock of an AI chat website with a brand style that will not look like a typical messaging app with message bubbles.
I really like the idea of putting the text input exactly where the content would be, so the assistant's output naturally appears below it.
I don't know if this will make its way into a production app, but at least it was cool to build this demo.
What do you think?
I really like the idea of putting the text input exactly where the content would be, so the assistant's output naturally appears below it.
I don't know if this will make its way into a production app, but at least it was cool to build this demo.
What do you think?
This media is not supported in your browser
VIEW IN TELEGRAM
Today at FoxChat:
- the website now auto-scrolls while the LLM generates the text
- the input box is now sticks to the bottom of the page when there's not enough space
- small UI improvements to help distinguish between messages and less "UI jumping"
- the website now auto-scrolls while the LLM generates the text
- the input box is now sticks to the bottom of the page when there's not enough space
- small UI improvements to help distinguish between messages and less "UI jumping"