Token Counter · o200k_base text count, estimates elsewhere

AI token counter

Paste a prompt to count its characters, words, and o200k_base tokens in your browser. You’ll also see a clearly labelled estimate for other model families and the input cost at each published API rate.

Exact for the pasted o200k_base text · API wrappers can add tokens Runs in your browser — nothing is uploaded

Your text Open a local file
TXT, Markdown, CSV, JSON, HTML, PDF, or DOCX · parsed in this browserChoose file

0 Tokens exact tokenizer loads when you type

0 Characters

0 Words

0 RAG chunks 800 tok

0 Document pages ≈ 500 words

≈ 0 tokens for model families without a browser tokenizer. The rough estimate adjusts for Arabic, CJK, and other scripts.

Input cost of this text

cheapest first · input price only

Loading pricing…

Cost = tokens × each model's input price from the pricing index. Output isn't counted here — for full monthly cost use the calculator. Self-hosted models with no API price are omitted. Prices verified June 2026; see methodology.

How token counting works here

Every model splits text into tokens with its own tokenizer, so one paragraph is not the same number of tokens everywhere. benchr is honest about which numbers it can prove:

o200k_base pasted text: benchr runs the encoding in your browser, so the tokenization of the raw text itself is exact. An API request can add message framing, tool definitions, images, or model-specific overhead, so this is not a guarantee of the billed input total.
Claude, Gemini, Grok, DeepSeek and others (estimate): these families don't ship a public browser tokenizer. benchr uses a rough, script-aware character estimate and labels it clearly — never as an exact provider count.

Word and character counts are exact for what you paste. The cost figure multiplies this text count or family estimate by the listed input price; it is a planning estimate, not an API invoice.

Also useful

→ Cost calculator — monthly cost across every model → Pricing index — input, output, and cached rates → Model rankings — capability and price together → How to cut your token bill

Frequently asked questions

Are the token counts exact?

The o200k_base count is exact for the raw text pasted into the box. API-billed input may differ because messages, tools, images, and model-specific wrappers add tokens. Claude, Gemini, Grok, and other families are labelled script-aware estimates.

Why are Claude and Gemini only estimates?

Anthropic and Google don't publish a client-side tokenizer the way OpenAI's encodings are openly available. Rather than show a fake-precise number, benchr labels those counts as approximate. Treat them as a guide, not a billing figure.

Is my text sent anywhere?

No. Counting runs entirely in your browser — the text you paste never leaves your device. benchr only fetches the public pricing file to compute cost.

How is the cost figure calculated?

It multiplies the token count by each model's published input price from the benchr model index (per million tokens). It is the input cost of the pasted text; for full monthly cost including output, use the cost calculator.