LLM Token Counter and Cost Estimator
Count tokens for GPT-5, Claude 4.7 and Gemini 2.5. Estimate request cost in real time.
LLM Token Counter and Cost Estimator
How it works
- 1.
Pick a model
Click a model — GPT-5, GPT-4.1, o3, o4-mini, Claude Opus 4.7, Gemini 2.5 Pro, Llama 4 and more.
- 2.
Paste your prompt
Paste your prompt or text. Token count and estimated cost update live. A slice of bread is about 3 tokens, a paragraph around 70.
- 3.
Estimate output tokens
Enter the expected response length to see total cost per request (input + output). Output tokens are 3-5× more expensive than input on most models.
FAQ
Are my prompts sent to OpenAI?+
No. Tokenization runs entirely in your browser using a local copy of the tokenizer.
How accurate is the count for Claude, Gemini and Llama?+
Exact for GPT models. For Claude, Gemini and Llama we use a close approximation (~4 chars per token). Typical deviation under 10%.
Are the prices current?+
Prices shown (USD / 1M tokens) reflect 2026 rates. Compare with the provider's official pricing page when needed.