LLM Context & Usage Meter

Estimate context-window usage and warn before LLM degradation.

Stay ahead of context limits in long AI conversations. Keep track of your AI conversation's context window before response quality starts to drop. LLM Context & Usage Meter runs directly in your browser to estimate token usage from your visible chat and displays a live gauge of your remaining context limit. Designed for heavy users of generative AI, this tool helps you manage long-running conversations without hitting sudden cut-offs or experiencing degraded, forgetful responses. Core Features • Live Context Meter: View your token usage via an in-page gauge and a convenient browser popup. • Proactive Alerts: Visual warning states alert you when context usage reaches degradation or critical levels. • Local Token Estimation: Estimates are based on the visible message text and specific model profiles, with options for manual model overrides when needed. • Summarize & Continue: Seamlessly transfer a heavy conversation into a fresh thread by generating a summary prompt to carry your context forward. • Usage Analytics: Access local conversation history and peak-token analytics to understand your chatting habits. • Smart Loading: Includes full-history loading capabilities for supported virtualized conversations. Supported Platforms The meter integrates with popular web interfaces such as ChatGPT, Claude, and Gemini. Privacy by Design Your chat data stays on your machine. Conversation text is processed entirely locally to estimate token usage and identify message roles. Raw conversation text is never uploaded to the publisher, sent to remote servers, stored as retrievable history, or included in telemetry. The extension bundles all necessary code locally and does not load executable scripts from remote URLs. Free vs. Pro The Free plan includes the standard meter and core local tracking for one selected platform. Upgrading to Pro unlocks tracking across all supported platforms, custom warning thresholds, additional appearance palettes, and the complete "Summarize & Continue" workflow. Pro checkout and subscriptions are managed securely via ExtensionPay. Note on Estimates: Token counts are local estimates based on visible text, not provider-authoritative measurements. Actual provider limits may vary as they can include hidden system prompts, tool calls, or metadata not visible in the browser.