Token counter
Count GPT-style BPE tokens with o200k, cl100k, p50k, p50k edit, and r50k tokenizers.
Prompt Info
Know the bill before the model runs.
Prompt Info is a browser-based LLM token counter and cost calculator for people who need to check how many tokens are in a prompt before a model request runs. Paste draft text, choose a model, compare OpenAI tokenizer families, estimate output size, and export a planning receipt for longer agent sessions.
By Hello.World Consulting. .
Cost calculator
Estimate input, output, and multi-turn agent costs before sending expensive model calls.
Prompt format lab
Compare the same prompt as TOON, JSON, YAML, XML, CSV, and compact JSON for payload planning.
Workbench
Price the request in the same order you build it.
Start with text, then select the model and tokenizer. Output uses a conservative auto estimate until you switch to a preset or custom number.
Step 01
Paste the exact text you plan to send. The receipt updates as you type.
19% of the 4,096 token fallback planning window
Token evidence
Inspect what the counter sees.
Paste a prompt to see token fragments and IDs.
Need a different payload shape?