Should this prompt leave your laptop? Building a local-first AI router and doing the honest math
Building a local-first AI router, and doing the honest math on what it saves. AI disclosure: I built this project with an AI coding assistant (Claude), which also helped draft this article. The benchmark ran on my own laptop, and I have checked every number, source and claim below against the code and the raw results, which are all public. Every AI feature you use, a chatbot, a writing assistant,…
GreenRoute is a Python service designed to determine whether an AI request should be handled locally on a user's device or sent to a cloud-based service. The tool aims to minimize energy consumption and carbon emissions associated with AI processing. Built by the author using an AI coding assistant, the system classifies requests into types such as Q&A, summary, rewrite, translation, code, analysis, long-form writing, and image generation.
Each type has a base complexity score that adjusts based on factors like input length and desired output length. The router then routes requests to either local models like Ollama or cloud services like Google Gemini, explaining the decision process. GreenRoute estimates energy usage based on Google's published figures for cloud processing and calculates potential savings by comparing local and cloud-based processing.
The tool's design prioritizes transparency, providing a "Why?" panel for each answer detailing the checks that led to the decision. Tests on 60 everyday prompts showed that 58% were kept on the device, with results that matched the author's labels and safe options sent to the cloud.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
