This new release (BrainSoup 1.31.0) is a major update which introduces the following changes:

• New code signing certificate: The application is now signed with a new code signing certificate issued by Certum. As every new code signing certificate, it needs to gain trust from the operating system, so Windows SmartScreen may display a warning during installation saying that the publisher is not well known. This is normal, and you can safely ignore it. The certificate will gain trust over time as more users install the application.
• New language models: Support for the new GPT-4.1-Nano, GPT-4.1-Mini and GPT-4.1 models has been added. These models are more cost-effective than their predecessors. They are also better at using tools, and have a huge context window of 1M tokens.
• Improved overall speed and cost-effectiveness: We now use the GPT-4.1-Nano as the base model for all internal orchestration prompts in BrainSoup, resulting in a faster, more robust and cost-effective solution.
• Updated Mistral models pricing: The Mistral LLMs are now up to 5 times cheaper than before, making them a very attractive option.
• open-codestral-mamba is now deprecated and has been removed. Codestral is now the recommended model for coding tasks.
• Fixed a compatibility issue with OpenRouter.
• Updated Doc-E's knowledge to reflect the latest changes in BrainSoup.

This new release (TidyTabs 1.22.2) is a minor update which introduces the following changes:

• TidyTabs is now compiled with the latest Microsoft compiler.
• Slightly improved performance.
• Minor bug fixes.

This new release (AquaSnap 1.24.1) is a minor update which introduces the following change:

• AquaSnap is now compiled with the latest Microsoft compiler.
• Improved Windows 11 compatibility.
• Minor bug fixes.

This new release (BrainSoup 1.30.0) is a major update which introduces the following changes:

• Revamped Agent List: The agent list has been revamped to make it clearer and more efficient, especially when dealing with a large number of agents. Agents are now divided into two lists: participants and non-participants, the latter being collapsed when not needed. This change makes it easier to manage your agents and to focus on the most relevant ones.
• Updated the underlying gpt-4o model, which is now abble to output 16384 tokens instead of 4096, allowing for more complex and detailed responses.
• The "Stay active in background" icon has been changed to a robot icon, to better reflect that it is related to automating tasks.
• Updated Doc-E's knowledge to reflect the latest changes in BrainSoup.

This new release (BrainSoup 1.29.0) is a major update which introduces the following changes:

• Source Code Semantic Indexing: BrainSoup can index source code files semantically, allowing your agents to understand your codebase and answer any questions related to it. To do this, simply add your source code to the `Documents` folder or the Sandboxed File System.
• Improved RAG pipeline: Your agents are now much better at leveraging their knowledge to provide accurate and relevant responses.
• GitIgnore: New ability to ignore files and directories listed in '.gitignore' and '.BrainSoupIgnore' files. This feature is particularly useful when working with source code repositories, to prevent irrelevant files from being indexed.
• Better context window management: Adaptative algorithms are now used to get the most out of the available agent's context window.
• System and hidden files, as well as files and directories whose names begin with a dot ('.'), are now hidden to agents and excluded from indexing.
• The `FileSystem` context has been improved to make it able to handle very large file trees more efficiently, whithout consuming too much of the agent's working memory.
• Fixed a bug that reduced the ability of agents to remember past conversations.
• Updated Doc-E's knowledge to reflect the latest changes in BrainSoup.

This new release (BrainSoup 1.28.0) is a major update which introduces the following changes:

• Codestral Mamba!: This is a new model by Mistral has a huge 256K token context window, and is very cost-effective.
• GPT-4o pricing update: We are now using the experimental GPT-4o-2024-08-06 model instead of GPT-4o, which reduces costs by half! According to our tests, the new model is just as performant as the previous one, if not better. But please let us know if you experience any issues.
• Mistral Large pricing update: This powerful model is now 25% cheaper to run.
• Improved multi-agent experience: We've made some improvements to the multi-agent chat orchestration system to make conversations more natural and productive.
• Enhanced vision capabilities: The 'ImageToText', 'Camera', and 'Screenshot' tools are now able to answer very specific questions about the image.
• Optimized Doc-E's settings to improve its performance.
• Fixed a bug that caused BrainSoup to sometimes fail to connect to the Ollama server.
• Fixed a bug that caused the shortcuts to the sandboxed file system to be incorrect when a custom location was used.

This new release (BrainSoup 1.27.1) is a minor update which introduces the following changes:

• A database backup is now created after each successful start, to prevent data loss in worst-case scenarios.
• GPT-4o-Mini is now the default LLM for all new agents.
• Improved the parsing of Markdown lists in the chat.
• Fixed a bug that caused the authentication token not to be refreshed immediately when the computer wakes up.
• Fixed a bug that caused the application to sometimes crash when changing chat room.
• Updated and consolidated Doc-E's knowledge.

This new release (BrainSoup 1.27.0) is a major update which introduces the following changes:

• GPT-4o mini! We've added support for the impressive new model GPT-4o-Mini. This model is smarter than GPT-3.5-Turbo, has a huge 128K token context window (vs 16K for GPT-3.5-Turbo), and is 3 times cheaper to run! We strongly recommend upgrading all your GPT-3.5-based agents to GPT-4o-Mini for massive savings and improved performance.
• Speech to text input! We've added a new feature that lets you speak to your agents by voice. This function is available via a button next to the input field, or via the keyboard shortcut 'Ctrl+S'. It is based on the OpenAI Whisper model, which is highly accurate and takes into account the context of the conversation.
• Mistral Nemo: We've added support for the new open-mistral-nemo model, a very interesting uncensored 12B LLM with a very large context window of 128K tokens, which is smarter than open-mixtral-7B, and only 20% more expensive to run.
• Improved overall cost-effectiveness: We now use GPT-4o-Mini as the base model for all internal orchestration prompts in BrainSoup, resulting in a more robust and cost-effective solution.
• Improved document indexing efficiency: This is a consequence of using GPT-4o-Mini for internal knowledge extraction prompts.
• More reliable WebBrowser tool: We have fixed a problem that sometimes prevented web searches from producing results.
• Better looking login screen.
• Fixed a bug that caused the application to sometimes crash when logging out.

This new release (BrainSoup 1.26.0) is a major update which introduces the following changes:

• New Email Custom Tool: This new tool allows your agents to send emails, with the ability to attach files, add cc/bcc recipients, and more. We have implemented this tool as a Custom Tool written in PowerShell, so you can see how easy it is to create your own tools for BrainSoup. To enable this tool, you just need to set your SMTP server settings in the 'Tools\Custom\Email\config.ps1' file.

This new release (BrainSoup 1.25.2) is a minor update which introduces the following changes:

• Fixed an issue causing some files not to be indexed.
• Fixed an issue preventing some scripting languages from being executed by the Hook manager.
• Updated Doc-E's knowledge about Hooks and how to use them.

This new release (BrainSoup 1.25.1) is a minor update which introduces the following changes:

• Fixed an issue where the application would crash at startup for some users.

This new release (BrainSoup 1.25.0) is a major update which introduces the following changes:

• New FileDownloader tool: This new tool allows your agents to download files from the internet.
• Improved ImageToText tool: The 'ImageToText' tool can now handle online images.
• Improved YouTube tool: Your agent can now also retrieve the title, author, and thumbnail URL of a YouTube video, in addition to the full transcript.
• Improved chat history limit: The pinned messages now never expire even if you set a limit on a chat's history, making sure agents never forget important information.
• Updated Doc-E's knowledge to include the latest changes.
• Fixed a bug causing sevaral file-based tools to stop working since the last release, sorry about that! 😅