IA de Guardia ES Follow

Series / AI at Home

Local and private

AI at Home

Models that run on your computer or your phone. With a local model they work without internet and your data does not leave the machine; privacy depends on the tool and on not connecting it to a remote provider.

Watch on TikTok (in Spanish) Watch on YouTube (in Spanish)
Black mini PC with a lime LED next to a laptop showing a local AI chat, on a wooden desk

Can I use AI on my own computer without sending my data to anyone?

That is the question this series answers. Each video comes with a fact sheet: what it is, the key fact, how to start and my verdict. How the fact sheet works.

Before you start.

Who it is for

For anyone who wants to use AI models without internet and without sending their data to anyone.

What to check

Look at how much memory and what graphics card each model needs, and its license: it is almost never the same as the program that runs it. Confirm the tool uses a local model and not a remote provider: several allow both. See the full checklist.

The 16 repos that have already appeared in this series.

3 of 16 already have their own guide (“Guide” label): license, requirements, installation and warnings. The rest open their project on GitHub. The license label comes from GitHub and each repo’s LICENSE file; each repo’s warning is still in its video (in Spanish).

  1. KoboldCpp Run local language models with a single executable. GitHub ↗ (opens GitHub in a new tab)

    License: AGPL-3.0

  2. GPT4All Chat with models that run on your computer. GitHub ↗ (opens GitHub in a new tab)

    License: MIT

  3. AnythingLLM Ask your documents questions with a local or cloud model. Guide (opens the guide)

    License: MIT Official project (opens GitHub in a new tab) Calculator

  4. Open WebUI Chat interface for your local models. Guide (opens the guide)

    License: Open WebUI License (BSD-3 with a branding clause) Official project (opens GitHub in a new tab) Calculator

  5. Piper Fast synthetic voice that runs without internet. GitHub ↗ (opens GitHub in a new tab)

    License: GPL-3.0

  6. llamafile A single file that runs a language model. GitHub ↗ (opens GitHub in a new tab)

    License: Apache-2.0

  7. stable-diffusion.cpp Generate images with Stable Diffusion in C++. GitHub ↗ (opens GitHub in a new tab)

    License: MIT

  8. whisper.cpp Transcribe audio without sending it to any cloud. Guide (opens the guide)

    License: MIT Official project (opens GitHub in a new tab) Calculator

  9. faster-whisper A faster Whisper for transcription. GitHub ↗ (opens GitHub in a new tab)

    License: MIT

  10. Bark Text to speech and sounds with AI. GitHub ↗ (opens GitHub in a new tab)

    License: MIT Calculator

  11. AudioCraft Generate music and audio with AI. GitHub ↗ (opens GitHub in a new tab)

    License: MIT

  12. InvokeAI Image generator with an interface built for real work. GitHub ↗ (opens GitHub in a new tab)

    License: Apache-2.0 Calculator

  13. Fooocus AI image generation without so many settings. GitHub ↗ (opens GitHub in a new tab)

    License: GPL-3.0 Calculator

  14. Krita AI Diffusion Generate AI images inside Krita. GitHub ↗ (opens GitHub in a new tab)

    License: GPL-3.0

  15. Jan A local alternative to ChatGPT. GitHub ↗ (opens GitHub in a new tab)

    License: Apache-2.0

  16. ComfyUI Node workflows for generating AI images. GitHub ↗ (opens GitHub in a new tab)

    License: GPL-3.0 Calculator

Search in the directory

← Previous seriesStop Paying