Tokens
Text is converted into tokens
The Context Window
Context Window
The context window is a fixed-size box
Free
16k
Go
32k
Plus
32k
Pro
128k
But your subscription tier decides how much you actually get
Multi-turn Conversation
Context Window
Every message takes up space in the box
What Happens When It's Full?
Older messages
silently removed
Middle of conversation
summarised
Older messages
summarised
Session limit reached.
Start a new chat.
Chat ends entirely
Providers handle this invisibly — and differently
Hidden Instructions
Context Window
Root
Cannot be overridden. Prevents catastrophic harm.
overrides ↓
System
Set by provider. Varies by surface & user type.
overrides ↓
Developer
API-level instructions from app builders.
These layers are already inside the box before you type
Long-Term Memory
Context Window
Past chats and stored user data pre-fill the box
Document Uploads
60k tokens
Context · 110k limit
Under 110k tokens — full document included
200k tokens
first 110k remaining 90k
Context Window
🔍
Private Search Index
Over 110k — the remainder is only searchable, not seen
What's Left?
Plus Plan · 32k tokens
Reasoning Tokens
Context Window
Reasoning tokens: temporary, invisible, discarded
The model is never operating on
what you think it is.
Should we demand more transparency?