LLM Context Anatomy

Token Budget & Context Window Visualizer

System Prompt2,000 tok
Multimodal Input0 tok
Images0 files
Audio (Speech)0 mins
Video0 mins
Chat History5,000 tok

10 turns

MCP Tools0 tok
RAG Documents0 tok

0 chunks

Context Window Usage

7,150/ 128,000 tokens
5.6%
Capacity Used
System
System Prompt
Images/Audio/Video
History
Chat History
Tool Schemas
Retrieved Context
Current Input

Scale Analogy

Your Context0.0084 Bibles
Model Capacity0.15 Bibles

Note on Media Tokens: Video quality matters! 4K video consumes ~3.5x more tokens per minute than Low Res because models sample more frames or use higher resolution grids to "see" the details.