Back to Home
Inside an LLM
Context Window
A context window is the maximum amount of text an AI can hold in memory at one time. When the conversation grows longer than the window, earlier messages get cut off, this is exactly why long chats sometimes seem like the AI "forgets" what you said earlier.
This is a simple learning demo, not an exact model limit calculator.
Interactive Playground
80
AI Reading Window
token
Inside context
token
Outside context (cut off)
Type something to see tokens appear…
Context Usage
0 / 80 tokens
0
80
Statistics
0
Total Tokens
80
Context Limit
0
Inside Context
0
Outside Context
0%
Context Used
How It Works
📝
Text
→
🔤
Tokens
→
🪟
Context Window
→
🤖
AI Reads
🎓
Beginner Tips
2 tips
Tokens outside the window are completely invisible to the model, it can't summarise or refer to what was cut off.
Context windows are in tokens, not words. 128K tokens ≈ 90,000 words, roughly one full novel.
💡
Key Takeaway
AI can only work with the text that fits inside its context window. Tokens that overflow are invisible to the model.