Private AI utility
Context Window Calculator
See how much of each model’s context window your prompt uses. Everything runs in your browser; your prompt is never uploaded.
GPT-50.00% · 400,000 tokens
GPT-4.10.00% · 1,047,576 tokens
Claude 4 Opus0.00% · 200,000 tokens
Claude Sonnet0.00% · 200,000 tokens
Gemini 2.50.00% · 1,048,576 tokens
Llama 40.00% · 1,000,000 tokens
Mistral Large0.00% · 128,000 tokens
DeepSeek V30.00% · 128,000 tokens
How it works Context Window Calculator
Your estimated token count is divided by each configured context window. The warning state appears when use reaches 80%, leaving space for system prompts and output.
Example
Paste a long retrieval context to see its share of each model’s maximum context window.
Frequently asked questions
- Why leave unused context?
- Chat history, tool responses, and the model output all consume the same context window.
- What happens at 100%?
- The request may fail, truncate text, or leave no useful room for a response.