About
☰
All Tools
Categories
Blog About

AI Token Counter - Count Tokens for Prompts & Chat Payloads

Pick a model, paste text or chat JSON, and get instant token counts (and cost estimates) to avoid context-limit errors and control spend.

🟢 Runs locally · no uploads

AI Token Counter

Count tokens for prompts, messages JSON, and Chat Completions payloads

Settings

Input

Input161 bytes · 161 chars · 26 words
Rewrite this prompt to be clearer and shorter, without changing meaning: "Please go over the following JSON and extract the relevant fields for an API request."

Output

Tokens (input)
0
Chars
161
3 lines
Bytes
161
UTF-8 estimate
Cost (est.)
N/A
In: N/A · Out: N/A

Related tools

Show more

JSON Formatter & Validator

Stop squinting at minified JSON! Format and validate JSON instantly with automatic error detection, syntax highlighting, and pretty printing. Catches missing quotes, trailing commas, and invalid syntax. Works with JSON and JSONC. Perfect for debugging API responses and config files. Free, works offline, unlimited usage.

JSON Path Extractor - Get, Search & Extract Values from JSON

Get specific values from JSON data instantly! Search through nested objects, extract array elements, and retrieve data using powerful JSONPath expressions. Perfect for getting exactly what you need from complex JSON structures.

AI CLI Assistant

Authoritative AI CLI reference covering Claude Code, Ollama, OpenAI, Vertex AI Gemini, Cursor, GitHub Copilot, and aider with syntax, flags, and safety notes.

Word Counter - Free Online Word & Character Counter

Stop manually counting words in documents! Instantly count words, characters (with and without spaces), sentences, paragraphs, and estimate reading time. Perfect for essays, articles, social media posts, and SEO content. Free, works offline, unlimited word counting.

Text Analysis Tool - Comprehensive Word Counter & Text Analytics

Comprehensive text analysis tool combining word counting, character counting, readability assessment, sentiment analysis, keyword extraction, entity recognition, and quality metrics. Perfect for writers, students, content creators, and researchers who need detailed insights into their text.
Show more

Git Diff Visualizer

Advanced Git diff analysis with visual representation, code quality assessment, security insights, and comprehensive change impact analysis

JSON Flattener

Flatten nested JSON objects to flat key-value pairs or unflatten back to nested structure with customizable delimiters and array handling

JSON Workspace - Format, Extract & Transform

Complete JSON toolkit in one unified workspace. Format and beautify JSON, extract specific values with JSONPath queries, flatten nested structures, and transform data - all without switching tools. Perfect for developers working with complex JSON data.

Configuration File Validator

Comprehensive validation for configuration files including JSON, YAML, TOML, XML, Dockerfile, Nginx, Apache configs with syntax, security, and best practices checking

Validate YAML - YAML Validator & Syntax Checker

Validate YAML syntax online with comprehensive error checking, indentation validation, and structure analysis. Free YAML validator for configuration files, Docker Compose, Kubernetes manifests, and more.
› About this tool · FAQ

Count tokens for AI prompts and OpenAI-style chat payloads before you send them. Paste plain text, a messages array, or a full Chat Completions JSON request and instantly see token counts for popular models with cost estimates. Everything runs locally in your browser.

What is a token?

A token is a chunk of text used by language models. Tokens are not the same as words: some words split into multiple tokens, and punctuation/whitespace can also count.

Does this token counter support OpenAI chat payloads?

Yes. You can paste either a messages array (role/content objects) or a full Chat Completions request JSON with a messages field.

Why do token counts differ between models?

Different model families use different token encodings, which can change how text is split into tokens. This tool counts tokens using model-aware encodings.

Is my prompt uploaded anywhere?

No. All counting runs locally in your browser. Your text is not sent to a server.