Context Window Visualizer

See how much of a model's context window your prompt actually uses, and how much room is left for the response.

tokens
~0 / 127,000 usable tokens0.0%
127,000 tokens of headroom remaining before hitting the reserved-output boundary.
Full context window: 128,000 tokensReserved for output: 1,000 tokens

Token counts are an approximation. Leave headroom beyond the model's advertised limit — real tokenizers vary and some providers count formatting/tool-call overhead against the same budget.

About the Context Window Visualizer

Context window limits are shared between the prompt, any retrieved context, and the model's response — a long RAG prompt can silently crowd out the room a model needs to answer. This tool shows what percentage of a model's context window your current text uses, after reserving space for the expected output.

Key Features

  • Covers context windows from 16K (GPT-3.5) up to 2M tokens (Gemini 1.5 Pro).
  • Reserve a token budget for the response and see remaining headroom.
  • 100% client-side — your text never leaves the browser.