# TokenCount

> Count words, characters, and estimate GPT tokens with readability. Use when tracking length, checking budgets, comparing complexity.

- Skill: `bytesagain/tokencount` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add bytesagain/tokencount`
- Raw SKILL.md: https://api.skillmd.com/api/skills/bytesagain/tokencount/raw
- Safety review: pending (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: bytesagain (https://skillmd.com/u/bytesagain)
- Updated: 2026-09-09
- Page: https://skillmd.com/skills/bytesagain/tokencount

---


# TokenCount

A real text and token counter for the terminal. Count words, lines, characters, and sentences. Estimate LLM token usage with cost projections. Analyze word frequency and compare files side by side.

## Commands

| Command | Description |
|---------|-------------|
| `tokencount count <file\|text>` | Count words, lines, characters, sentences, paragraphs, avg word length, and reading time. Works with files or inline text |
| `tokencount tokens <file>` | Estimate LLM token count using 3 methods (chars÷4, words×1.33, bytes÷3.5), shows cost estimates for GPT-4 class models and context window usage bars |
| `tokencount freq <file>` | Full word frequency analysis — ranked table with counts, percentages, bar chart, and vocabulary richness score |
| `tokencount top <file> [n]` | Show top N most common words (default: 20) |
| `tokencount diff <file1> <file2>` | Compare two files — side-by-side word/line/char/token counts, unique words in each, common vocabulary |

## Requirements

- Standard Unix tools: `wc`, `awk`, `sort`, `tr`, `grep`, `comm`

## Examples

```bash
# Count everything in a file
tokencount count README.md

# Count inline text
tokencount count "Hello world, this is a test."

# Estimate tokens and costs
tokencount tokens article.txt

# Word frequency analysis
tokencount freq novel.txt

# Top 10 words
tokencount top essay.txt 10

# Compare two drafts
tokencount diff draft-v1.txt draft-v2.txt
```