ChatGPT Plugins are tools specifically designed for language models with safety as a core principle, enabling ChatGPT to access up-to-date information, perform computations, or utilize third-party services. Plugin developers, who have been invited off the waitlist, can build plugins for ChatGPT using the provided documentation. Enabled plugins are listed in the prompt shown to the language model, along with instructions on how to use each plugin. Plugins allow language models to access real-world and recent data, acting as their "eyes and ears." They provide access to information that may be too recent, personal, or specific to be included in the training data. Plugins can also enable language models to perform safe and constrained actions on users' behalf, enhancing the system's overall usefulness. Plugins address challenges associated with large language models, such as hallucinations, staying up-to-date with recent events, and accessing proprietary information sources (with permission). By integrating explicit access to external data sources, language models can strengthen their responses with evidence-based references. The initial set of plugins has been developed by companies like Expedia, FiscalNote, Instacart, KAYAK, Klarna, Milo, OpenTable, Shopify, Slack, Speak, Wolfram, and Zapier.