ComfyUI · video · images · local LLMs

Local AI,
step by step.

Learn ComfyUI, AI image and video generation, and local LLMs through workflows designed to be understood—not just copied. Every guide treats 8 GB VRAM as a real design target, not an afterthought.

A starting point for each model

Start with the setup.
Then try something new.

Follow a series from the beginning, or go straight to the task you need. Each guide includes the original video, chapter links, and hardware context.

Browse all tutorials

Consumer hardware, explained

8 GB VRAM,
with the full picture.

The local setup demonstrations use this laptop configuration. GPU memory is only part of the story: system RAM, offloading, resolution, and clip length matter too.

How Tensor Alchemist approaches local AI
Demonstrated GPU
NVIDIA RTX 5060 Laptop
GPU memory
8 GB VRAM
System memory
32 GB RAM

These are the tutorial hardware details, not minimum requirements or a promise that every workload fits every 8 GB GPU. Check the notes in the guide you choose.

Recently published

The latest
walkthroughs.

New guides from the channel, with the key steps organized for easy reference.

All tutorials
  1. MiniMax H3

    MiniMax H3: Two-Pass Latent Upscaling

    Follow the Turbo LoRA workflow from a coarse first pass through learned latent upscaling and a second detail pass, with text-to-video and first/last-frame modes.

    8:28 video · Chapter links included
  2. MiniMax H3

    MiniMax H3: Tracked Video Inpainting

    Track an object with SAM 3.1, clean its mask, and regenerate a cropped region before blending it back into the source video.

    9:33 video · Chapter links included
  3. LTX 2.5

    LTX 2.5: Direct the Shot

    Arrange first and last frames, additional image guides, timed prompts, and video references in Director 2.0 to give a shot a clearer structure.

    8:11 video · Chapter links included