WritingTool can be supported by an AI. The application is optimized for the use of localAI, which enables local AI installation on a PC or laptop of today's standard.

Note: The performance of a local AI ​​application depends very much on the performance of the processors and the available memory.

A short installation guide is provided here. It only covers the simplest case. For more in-depth options, e.g. the more powerful use of GPUs from modern graphics cards, please refer to the original documentation (see https://localai.io/).

The simplest option is to install a complete package using docker.

Docker is a technology that runs an application encapsulated in a container. It is available for all common PC or notebook platforms today. It is an open-source product and free for private use and small businesses.

First, the Docker Desktop must be installed. The software and corresponding instructions can be found here:

Note: Under Linux, installing a Docker Engine is also sufficient (https://docs.docker.com/engine/install/

Next, a command window will open in the respective operating system:

  • Windows: Press the Windows key, enter cmd and click OK (Return). A console window will open.
  • Linux: Open terminal window
  • Mac: Open terminal window

The following line is then entered in the window and concluded with Return:

  • docker run -p 8080:8080 --name local-ai -ti localai/localai:latest-aio-cpu

The localAI software and a basic set of AI images (> 10 GB) will be downloaded. This takes a while.

After successful installation, you can call up the AI ​​in a browser at the following address:

Additional models can also be loaded via the browser interface. This is necessary, for example, if you want to use the TextToSpeech (TTS) functionality in your own language or with a voice of your own choosing.

The default settings in the AI ​​options of WritingTool correspond to the standard installation described above.