Estimate token counts and API costs for popular LLMs
0
Estimated Tokens
0
Characters
0
Words
0
Lines
Estimated API Cost (for this text as input)
How it works: Token counts are estimated using a heuristic tokenizer that approximates BPE (byte-pair encoding) used by most LLMs.
Actual counts may vary ±10% depending on the model's specific tokenizer. For exact counts, use each provider's official tokenizer.
Tip: Most models count ~1 token per 4 characters of English text, but code, non-English text, and special characters use more tokens.