AI Token Tracker

Track token usage across Claude, ChatGPT, Gemini, and more — automatically.

AI Token Tracker estimates the cost of your AI conversations as you have them. Every time you chat with an AI, tokens are consumed, compute is billed, and water is used to cool the data center. None of that is visible by default. This extension makes it visible. Open the side panel to see a live breakdown of your current conversation: an estimated token count, an approximate cost based on current API rates for your chosen model, and a water footprint estimate based on published research. Everything is calculated locally on your device — nothing is sent to any server. WHAT IT ESTIMATES Token count for each message and the full conversation, cost based on current API pricing for your chosen model, and water usage derived from peer-reviewed data on AI data center consumption. HOW IT WORKS The extension reads the text already visible on your screen and counts tokens locally using bundled tokenizer files. No conversation content is transmitted anywhere. The only optional network request is to Anthropic's official token-counting API, and only if you choose to enter an API key. SUPPORTED PLATFORMS Works on the most widely used AI chat interfaces, including Claude, ChatGPT, Gemini, Grok, and Perplexity. See the full list of supported platforms in the screenshots below. PRIVACY No conversation data leaves your device. No analytics. No telemetry. No ads. All pricing data, fonts, and tokenizer files are bundled inside the extension — nothing is loaded from external servers at runtime. The full privacy policy is available via the link below. A NOTE ON ACCURACY Cost and token figures are estimates. Water estimates are order-of-magnitude figures based on academic research; actual consumption varies significantly by data center and cooling method.