OpenAI Expands Responses API With MCP Support and Image Generation
OpenAI adds remote MCP server support, native image generation, and upgraded file search to its Responses API to help developers build advanced agentic applications.
OpenAI introduces major updates to its Responses API to help developers build intelligent, action-oriented AI agents. The latest enhancements include support for remote Model Context Protocol (MCP) servers, native image generation through GPT-4o, and upgraded file search capabilities. These tools allow developers to connect external services like Stripe and Shopify directly to OpenAI models with minimal code.
Since its launch in March 2025 alongside the open-source Agents SDK, the Responses API processes trillions of tokens for diverse use cases ranging from market research to financial analysis. Popular applications built on this API include Zencoder's coding agent, Revi's market intelligence assistant, and MagicSchool's educational platform. The API essentially provides third-party developers with the same underlying technology that powers OpenAI's own ChatGPT, Deep Research, and Operator services.
This update represents a significant step toward OpenAI's goal of deprecating the Chat Completions API by mid-2026 in favor of this unified toolkit. By combining elements of previous APIs into a single interface, OpenAI gives enterprises a frictionless way to create production-ready, domain-specific AI agents. The addition of remote MCP support and Code Interpreter tools further eliminates the need for complex orchestration logic in autonomous workflows.