Braina 2.1 – Advanced AI Automation, Gemini 1.5 Pro/Flash, OpenAI GPT-4 Omni

New features and improvements made in Braina version 2.1 are as follows:

  1. New Advanced AI Chat engines/ LLMs: OpenAI GPT-4 Omni, Gemini Pro 1.5 & Gemini Pro 1.0

    GPT-4 Omni is OpenAI’s most advanced multimodal model that’s faster and cheaper than GPT-4 Turbo with stronger vision and inbuilt voice capabilities. The model has 128K context and an October 2023 knowledge cutoff.

    Gemini Pro 1.5 is Google’s next-generation model with a breakthrough 2 million context window. (around 140K words)

    Gemini Flash 1.5 is faster and cheaper model with 1 million context window (around 70K words)

    You can find more information about the advanced AI credits here: https://braina.me/shop/

  2. Advanced AI Automation (Beta) – FREE for a Week!

    You can now use Braina Swift and Braina Pinnacle LLMs to automate actions on your computer! At present, only keyboard automation is available. Support for mouse automation and the Large Action Model (LAM) will be introduced soon. Advanced AI Automation feature will be free for all users (Lite or Pro) till 22nd May 2024.

    Just prepend “Use keyboard” before the task to be performed and Braina Swift or Braina Pinnacle will automate keypresses for you. For complex tasks that require multiple steps, you can specify a series of keyboard actions to perform and Braina will do it for you. Some examples are shared below.

    Tip: Enable “LLM Mode” for faster execution and avoid conflicts with inbuilt commands.

    Press Keyboard shortcuts

    Press Ctrl+Enter
    Press Alt plus F4

    Basic automation prompts examples

    Use keyboard to close window
    Use keyboard to open New York times website in chrome browser
    Please open notepad and write a motivational quote using keyboard
    Use keyboard to list all jpg files on my desktop
    Use keyboard and PowerShell to list all files in Downloads folder that were downloaded today
    Use keyboard to open an interesting website to surprise me

    Complex automation prompt example

    If you want to perform complex tasks via automation, you will require to mention the steps that should be followed. Here is an example prompt to automate posting a tweet on Twitter (X).

    Use keyboard to post tweet about my good experience with Braina AI on twitter with appropriate hashtags.
    Steps:
    1) Open https://twitter.com/compose/post in chrome browser. Wait for 4 seconds.
    2) Write the post. Also add Braina’s website link: https://braina.ai . Add a space at the end. Wait for 2 seconds.
    3) Press Ctrl+Enter . It is the shortcut to post the tweet.

    Note: You may adjust the wait period as per your system responsiveness. If you have a slow system, you might need to change wait period of 4 seconds in the first step to 7 or more seconds. It is assumed that the user is already logged in his/her Twitter account.

    Create custom commands with parameter for Automation (Pro version required)

    You can create custom commands (or custom prompts) with parameters for to make advanced AI automation easier. For example, we can create a custom prompt “Post tweet about <x>” if we don’t want to provide all the steps each time we want to post tweet on different topics. Follow steps mentioned below to create custom command with parameter.

    1. Go to main menu -> Custom commands/prompts or just press Ctrl+Alt+C (or Ctrl+Shift+C) to open custom commands/prompts dialog.
    2. Click on the “+” icon to create new custom command.
    3. Enter “Post tweet about <x>” (without quotes) in command name field and click on Ok.
    4. Enter following automation prompt in “enter the command that will be executed” filed:

      Use keyboard to post tweet about <x> on twitter with appropriate hashtags.
      Steps:
      1) Open https://twitter.com/compose/post in chrome browser. Wait for 4 seconds.
      2) Write the post. Add a space at the end. Wait for 2 seconds.
      3) press Ctrl+Enter . It is the shortcut to post the tweet.

    5. Done. Now you can just speak “Post tweet about <topic>” and watch Braina automating tweet posting for you! Here <topic> is the topic that you want to post about. e.g. Post tweet about latest AI developments

    Tip: Once you’ve created the custom command, you can attach a photo (Ctrl+I) and simply say or type “Post tweet about this image.” Braina will then generate and post a relevant tweet about the image. If you prefer to review and approve the tweet before it goes live, omit the third step from the custom command. This way, Braina will only open Twitter and draft the tweet, allowing you to review it first. After confirming the tweet, you can post it manually or instruct Braina to press Ctrl + Enter while Twitter is in focus.

    Note: Features like image/file/webpage attachments, Braina’s persistent memory, web access all work with advanced AI automation.

    Important: Please note that Advanced AI automation is currently in Beta and it may not work sometimes or malfunction. Please do not use it for sensitive tasks such as deleting files. Currently, only English language is supported for advanced AI automation.

  3. Use your own OpenAI API key

    Now you can use your own OpenAI API key instead of advanced AI credits for using GPT-3.5 Turbo, GPT-4 Omni and GPT-4 Turbo LLMs in Braina. You will require Pro version to use your own OpenAI API key. Go to Settings -> AI engine and enter your API key under OpenAI API Key field and click Ok.

  4. Braina Swift & Braina Pinnacle update

    Braina Swift and Braina Pinnacle LLMs can now execute inbuilt commands in LLM mode for reminders, image creation, weather info, YouTube search etc., and they also now show related question suggestions. Some examples of inbuilt command calling is provided below:

    Reminders in LLM Mode

    Now you can ask Braina Swift or Braina Pinnacle to set reminders in more natural language and with multi-turn conversations. No need to follow rigid syntax.

    Reminders in LLM Mode

    AI Image creation using LLM Mode

    Image creation using Braina Swift/Pinnacle LLM can be fun and useful as the LLM can prompt the image generation model with detailed descriptions, allowing for the generation of highly customized and imaginative visuals based on specific concepts or themes. This process enhances image generation creativity. Here are some screenshots of images generated using LLM mode:

    Note: Using LLM mode for executing Built-in commands is in Beta and it may not work for every command or feature.

Leave a Reply

Your email address will not be published. Required fields are marked *