LLM Token & Cost Lens – Castov (VS Code Extension)

Real-time token counter & LLM cost estimator for Claude, GPT-4o, Gemini, and DeepSeek directly inside Visual Studio Code and Cursor IDE.

Features CodeLens file summaries, instant hover inspector, real-time status bar tracker, and an interactive cost comparison calculator with monthly production scale simulation.

Zero dependencies, instant local execution (<1ms), zero behavioral telemetry.

Powered by Castov.com.

✨ Features

1. 🔍 Hover Token & Cost Inspector

• Hover over any selected text, prompt string, or code block.
• Instantly view estimated token count, character count, word count, and a model cost breakdown table (Claude 3.5 Sonnet, GPT-4o, Gemini 2.5 Flash, DeepSeek-R1).

2. 📊 CodeLens File Summaries

• Shows an unobtrusive summary at Line 1 of your active file: 📊 File: 1.4k tokens (~$0.004 Claude • $0.0001 Gemini) | 0.7% of 200k
• One click opens the interactive Cost Calculator with the entire file preloaded.

3. ⚡ Dynamic Status Bar Tracker

• Status bar displays real-time token count: $(hubot) 1.2k tokens.
• When you highlight text, dynamically switches to show selection token count: $(hubot) Selection: 240 tokens (~$0.0007).

4. 🧮 Interactive Calculator & Monthly Production Simulator

• Paste any system prompt, RAG context, or user query.
• Real-time token counts as you type.
• Visual comparative cost bars across 6+ frontier models.
• Production Scale Simulator: See what your prompt will cost at 1,000, 10,000, 50,000, or 100,000 calls per day.

🤖 Supported Models & Calibrated Pricing (October 2026)

⌨️ Commands

📄 License

MIT © Castov