# Instruments ### What is an Instrument? On The Grid an **Instrument** is a standardized contract that defines what you're buying: a guaranteed minimum level of intelligence, latency, throughput, context window, and output length. * Instead of choosing a specific model from a specific provider, you choose an Instrument grade (like Text Prime for deep reasoning or Text Standard for fast responses), and any supplier model that meets or exceeds the specification can fulfill your request. * This is what makes inference fungible on The Grid - you trade a quality guarantee, not a model name. Each Instrument is governed by a **Specification**. It is a bundle of thresholds that all conforming models must meet, such as: * Performance score on [Artificial Analysis Intelligence Index](https://artificialanalysis.ai/evaluations/artificial-analysis-intelligence-index). * Throughput in tokens per second. * Time to first token. * Context window size. * Maximum output length. As long as a supplier model meets or exceeds the thresholds for an Instrument, they can deliver that Instrument. We manage our suppliers and monitor their performance to ensure that all suppliers meet these specifications. ### Current Instruments Today The Grid focuses on text to text Instruments, such as: * **Text Max**: for the most complex tasks * **Text Prime**: for high reasoning quality and long outputs * **Text Standard**: for low latency and high throughput [Read more about our Current Instruments and Specifications here.](https://thegrid.ai/docs/introduction/instruments-and-specifications/instrument-specifications-latest) {% hint style="info" %} Several more Instruments are on the near term roadmap as the platform scales. If you have suggestions or feedback on the Instruments we should support, please contact {% endhint %}