Models Pie

Overview
Models Pie is a comparison and ranking tool designed to help developers and AI users evaluate large language models (LLMs) based on specific performance and economic priorities. By aggregating data from BenchLM, OpenRouter, and Artificial Analysis, the platform allows users to rank models across three core dimensions: quality, cost, and speed.
The tool is intended for anyone looking to optimize their LLM selection process for specific workloads. Users can customize their requirements by adjusting the weights of these three factors to find the best-matching models for their needs, whether they prioritize high performance, cost-efficiency, or rapid throughput.
How to use it?
- Adjust the interactive pie chart sliders to set your desired balance between quality, cost, and speed.
- Select a preset configuration if you prefer a pre-defined trade-off like ‘Fastest’ or ‘Highest quality’.
- Review the ranked list of models, which displays specific metrics for quality scores, throughput, latency, and price per million tokens to make your final selection.
Features
Interactive weight adjustment, Real-time API pricing integration, Performance and latency rankings, Quality score benchmarking, Customizable model filtering, Preset comparison modes
