Arcade

Arcade (arcade.dev) is a developer platform that enables AI agents and large language models (LLMs) to interact with external tools and services. The platform provides infrastructure for tool-calling, allowing LLMs to delegate specialized tasks — such as calculations, cross-application workflows, and real-world actions — to dedicated tools rather than relying solely on the model's internal reasoning. Core capabilities and offerings include: - A tool-building and tool-calling framework that connects LLMs to external functions and APIs - A tool evaluation framework that allows developers to test and tune tools for model accuracy and reliability - Support for the Model Context Protocol (MCP), which serves as an integration layer for AI agents - Hosted toolkits and integrations enabling AI agents to perform tasks across third-party applications (e.g., reading documents, sending emails, writing to cloud services) - Arcade Labs, a public space for research, prototypes, and experiments The platform targets software developers and teams building AI-powered products who want to extend LLM capabilities beyond text generation. Arcade positions itself as infrastructure for the "agent stack," helping developers wire models to real-world actions efficiently and cost-effectively. Arcade supports multiple LLM providers (including OpenAI and Anthropic models) and publishes benchmarking and evaluation content to help developers understand model tool-calling performance. The company operates as a product-based SaaS platform with self-serve registration and accompanying developer documentation.