Real-time BPE Tokenization Analyzer
Prompt Tokenizer & Cost Estimator
Paste your prompt, code snippets, or system instructions to see exact characters, words, estimated token counts, and cost per 10k executions.
Load Preset:
Prompt & System Instruction Input
469 characters·65 words
≈ 124 tokens
Per-Execution Cost Breakdown
Real-time input token execution cost for this exact prompt across top foundational models.
GPT-4o (Omni)
$2.50 / 1M in
$0.00031
$3.10 / 10k calls
GPT-4o Mini
$0.15 / 1M in
$0.00002
$0.19 / 10k calls
o1 (Reasoning)
$15.00 / 1M in
$0.00186
$18.60 / 10k calls
o1-mini
$1.10 / 1M in
$0.00014
$1.36 / 10k calls
Claude 3.5 Sonnet
$3.00 / 1M in
$0.00037
$3.72 / 10k calls
Claude 3.5 Haiku
$0.80 / 1M in
$0.00010
$0.99 / 10k calls
What is a Token?
LLMs process language in chunks called tokens rather than full words. In English, 1 token is roughly 4 characters or 0.75 words. Punctuation, capitalization, emojis, and code syntax often consume additional tokens.
How to Compress Prompts to Reduce Costs
Eliminate conversational filler words in system prompts. Use concise bullet rules, abbreviated JSON schemas, and avoid repeating instructions already covered in few-shot examples.