GPT Token Counter
Paste a prompt to estimate its GPT tokens and what it costs on each OpenAI model.
~0
estimated tokens
0
words
0
characters
This is a character/word-based approximation, not the real tokenizer each provider uses internally — expect it to land within roughly 10–15% of the actual count for plain English text. For an exact count against Claude's own tokenizer, use the Messages API's
count_tokens endpoint.0 tokens
1.00× your input length — drag to model a shorter reply or a long generated response.
GPT-6 Astra
OpenAI
$0.00
$10/$50 per 1M
GPT-6.1 Sol
OpenAI
$0.00
$2/$10 per 1M
GPT-6.1 Terra
OpenAI
$0.00
$1/$5 per 1M
GPT-6 Luna
OpenAI
$0.00
$0.1/$0.5 per 1M
Claude Fable 5.1
Anthropic
$0.00
$10/$50 per 1M
Claude Opus 5.5
Anthropic
$0.00
$4/$20 per 1M
Claude Sonnet 5.5
Anthropic
$0.00
$2/$10 per 1M
Claude Haiku 5.5
Anthropic
$0.00
$0.1/$0.5 per 1M
Gemini 4 Argon
Google
$0.00
$2/$10 per 1M
Gemini 3.8 Flash
Google
$0.00
$0.75/$3.75 per 1M
Gemini 3.1 Pro (preview)
Google
$0.00
$2/$12 per 1M
Gemini 3.5 Flash-Lite
Google
$0.00
$0.3/$2.5 per 1M
Standard list prices from the Anthropic, OpenAI and Google pricing pages, checked 2026-10-31. For a monthly budget with caching and batch discounts, use the LLM API Cost Calculator.
OpenAI prices per 1M tokens
| Model | Input | Cached input | Output |
|---|---|---|---|
| GPT-6 Astra | $10 | $1 | $50 |
| GPT-6.1 Sol | $2 | $0.1 | $10 |
| GPT-6.1 Terra | $1 | $0.05 | $5 |
| GPT-6 Luna | $0.1 | $0.01 | $0.5 |
Standard rates from OpenAI's pricing page, checked 2026-10-31. Batch requests cost half. Estimate a monthly bill →
Counting GPT tokens
For an exact count, use OpenAI's open-source tiktoken library with your model's encoding, or read usage.input_tokens from an API response.
- →OpenAI publishes its tokenizers in the open-source tiktoken library, so exact offline counts are possible in Python and JavaScript.
- →Reasoning tokens are billed as output tokens even though you don't see them — budget for them on top of the visible answer.
- →GPT-6 prompts with more than 272K input tokens pay 2× input and 1.5× output for the whole request.
Frequently asked
How many tokens is a word for GPT models?
About 0.75 words per token for ordinary English, or roughly 4 characters per token. Code, numbers and other languages use more tokens per character.
Do reasoning tokens count toward the price?
Yes. Reasoning (thinking) tokens are billed at the output rate, so a short visible answer can still cost thousands of output tokens.
ToolNinja