Context window calculator — how full is 128k?

A raw token count does not say if you are safe inside a GPT-4o 128k context window (or 8k / 32k / 200k / 1M). This context window calculator counts your paste, shows fill %, and remaining tokens so you can trim before truncation. Switch OpenAI-style or approximate Claude/Gemini modes as needed.

Counting mode

Context window

Custom

Used0

Remaining128,000

Fill0.0%

Context window128,000

Good uses for this page

You are stuffing specs, logs, or transcripts into one message and need headroom.

You want to compare the same paste against 32k vs 128k windows.

You need a quick “percent full” signal before you hit send.

Related tools

Token counter Words ↔ tokens

Questions

Usually the listed window covers input plus output (and sometimes tools). Leave room for the answer — do not fill to 100%.

Estimates only. Not an official tokenizer from OpenAI, Anthropic, or Google. Verify with your provider before billing decisions.