Advertising 36 AI & Machine Learning 305 AI Assistants 551 AI Image 144 AI Video 79 AI Voice 126 Analytics 230 automation 1145 Cloud & Infrastructure 178 Communication 624 Copywriting 55 CRM 445 Database 228 Developer Tools 802 E-commerce 219 Email Marketing 180 File Storage 209 Finance 288 HR & Recruiting 58 IoT & Hardware 56 Media 204 Project Management 227 SaaS 484 Security 235 SEM & PPC 30 SEO 131 Social Media 224 Utilities 694 VoIP & Telephony 289
Back to Nodes

@jay0073/n8n-nodes-tokencalculator

v0.1.5
Last updated Sep 1, 2026

Privacy-first local token counting and LLM cost comparison for text, code files, and images in n8n

420 Weekly Downloads
420 Monthly Downloads
View on NPM GitHub Repository

Description

TokenCalculator.dev

Web GUI: https://tokencalculator.dev

A privacy-first, browser-based LLM token calculator for OpenAI, Anthropic Claude, Google Gemini, and DeepSeek. It measures text, readable documents, and supported images, then reports token usage, context-window fit, and API cost projections without uploading user content.

Features

  • Live token counts across four providers
  • Exact local BPE tokenization for supported OpenAI models
  • Deterministic Provider-Calibrated UTF-8 Projections for other tokenizer families
  • Client-side TXT, Markdown, code, CSV, JSON, PDF, and DOCX extraction
  • Provider image-token formulas where documented
  • Context-window overflow analysis and pricing projections
  • Static provider-specific SEO pages
  • No prompt or uploaded-file storage

    • Development

    npm install
npm run dev

    Quality checks:

    npm test
npm run build

    See CONTRIBUTING.md for repository structure, package-specific checks, pull request guidance, and release expectations. User-facing changes are recorded in CHANGELOG.md.

    Architecture

  • Astro static-site generation
  • TypeScript
  • Tailwind CSS v4
  • Web Worker tokenization
  • js-tiktoken for exact BPE counts on supported models
  • Models.dev pricing refresh with a bundled offline fallback

Provider pricing and tokenization behavior can change. Results identify whether they use exact BPE tokenization, a provider-published vision formula, or a deterministic provider-calibrated projection; provider API usage and invoices remain authoritative.

Companion package

The public @jay0073/tokencalculator-core package provides the shared token measurement, cost comparison, image formula, context-window, and workload APIs for Node.js and browser projects. Its source lives in packages/tokencalculator-core.

n8n community node

@jay0073/n8n-nodes-tokencalculator brings local text, PDF, DOCX, source-file, and image measurement into self-hosted n8n workflows. It supports token counting, single-model cost estimation, and multi-model comparisons without sending workflow content to TokenCalculator servers. See packages/n8n-nodes-tokencalculator.