← Back to all tools中文

Token Counter

Multi-model token counting and cost estimation for GPT, Claude, Llama, Qwen and DeepSeek

Input Text

Characters: 0Words (CJK counted per character): 0Lines: 0

Select model

Type or paste text to see live estimates.

Note: no official tokenizer is used; results are character-based estimates (~4 chars/token for English, ~1.5 chars/token for Chinese, adjusted per model tokenizer). Actual counts depend on each platform's tokenizer. Prices are approximate public list prices for reference only.