LLM Context Calculator
Check whether a prompt fits a model’s context window and how much room is left. Enter the window size, the tokens used by the system prompt and by documents or retrieved text, how many conversation turns you keep and their average size, and the room reserved for the answer. The calculator shows how the window is used, the tokens left or the overflow, and how many more turns or document pages would fit.
- Runs in your browser
- No sign-up
- Free to use
Optional: paste a word count to estimate tokens (about 1.33 tokens per English word).
Token counts depend on the model’s tokenizer; word-to-token conversion is an estimate. Use the Prompt Token Counter for a measured count.
How to use LLM Context Calculator
- Enter the model’s context window.
- Enter tokens for system prompt and documents.
- Set history and answer room.
- Read what is left.
LLM Context Calculator features
Full budget
System, documents, history, answer.
Overflow warning
When the prompt is too big.
What fits
Extra turns and pages.
Word estimate
Words to tokens.
Formula shown
Every result explains how it was calculated.
Private
Runs in your browser; nothing is sent or stored.
When to use LLM Context Calculator
- Designing RAG pipelines.
- Chatbot memory limits.
- Long-document summarisation.
- Choosing a model size.
LLM Context Calculator FAQ
Are the token counts exact?
Only if your inputs are measured. The word-to-token conversion is an estimate (about 1.33 tokens per English word).
Why reserve room for the answer?
Input and output share the window; a full prompt leaves no space for the reply.
What happens if I exceed the window?
The request fails or older text is cut, depending on the API.
How do I measure real tokens?
Use the Prompt Token Counter.
Budgeting a context window
Context windows are large but not free: every token costs money and time, and models pay less attention to the middle of very long prompts. Planning the budget keeps prompts focused.
Trimming conversation history and retrieving fewer, better passages are the usual levers.
A practical rule is to keep the prompt well below the limit – many teams aim for no more than half to three quarters of the window – so that long user messages, tool results or unexpected document sizes do not push the request over the edge in production.
Prices, limits and tokenizers differ between providers and change often. Enter the current values from your provider’s pricing and documentation pages, and re-check them before committing to a budget.
Measure real usage once you have it: average tokens per request in production are often different from the estimates used at the planning stage, especially once system prompts, retrieved context and conversation history are included.