Context Window
The context window is an AI model's short-term working memory. Measured in tokens (roughly ¾ of a word), it defines the maximum amount of text the model can read, hold in mind, and reference at one time during a single conversation.
Think of it like the size of your desk. A small 4,096-token context window holds about 3,000 words—enough for a casual chat, but paste in a long PDF or a couple of code files and the model immediately starts pushing your earliest instructions off the edge of the desk and forgetting them. Modern local models often support 32,000 to 128,000 tokens, letting you feed in entire documentation folders or long transcripts at once.
Just remember that advertised context limits and usable local limits are two different things. Even if a model supports a massive 128K context window, cranking the slider all the way up on a home rig consumes huge amounts of extra VRAM and slows down response times.
Looking for verified hardware rigs?
Check our real-time £/GB VRAM index to find systems optimized for local model inference.