Zoom has announced an expansion of its developer platform. The announcement, made at Zoomtopia 2026, introduces three major components: new Zoom AI Services APIs, expanded Model Context Protocol (MCP) support, and Platform Studio, a unified developer portal.

The initiative reflects a strategic shift in how Zoom approaches AI development. Rather than confining AI capabilities to its own applications, the company is opening access to the underlying technology that powers Zoom's products—allowing developers to build conversational voice, transcription, translation, and summarisation features into their own applications and integrate Zoom context into third-party AI assistants.

Zoom AI Services: Direct Access to Speech and Language Technology

Zoom AI Services gives developers programmatic access to speech, language, and voice infrastructure used within Zoom's own products. At Zoomtopia, the company announced 15 new services expanding the API portfolio.

Key new capabilities include:

  • Voice Agent APIs: Developers can combine speech recognition, large language model reasoning, and speech synthesis in a managed pipeline, with the option to bring their own LLM and connect it to Zoom's speech-to-text and text-to-speech technology.
  • Text-to-Speech APIs: Support for both synchronous and live-streaming applications.
  • Interpreter API: Combines transcription, translation, and speech synthesis for near-real-time speech-to-speech interpretation.
  • Speech Summarisation APIs: Converts audio directly into structured summaries and insights, with customisable outputs for specific workflows.
  • Expanded Scribe API: Now includes automatic language detection, text polishing, time-synchronised captions, and new 3B Pro High-Accuracy models for batch, fast, and live transcription.
  • Decisions API: Evaluates text, images, and other inputs against predefined criteria, returning structured results for classification, scoring, and conditional evaluation tasks.

According to Zoom, the APIs allow developers to add capabilities like transcription, translation, summarisation, and voice to their applications without needing to build and operate the underlying AI infrastructure themselves.

Model Context Protocol Expansion: AI Assistants Beyond Zoom

Zoom is taking what the company describes as a "headless" approach, making Zoom intelligence and capabilities available within the AI environments where users already work.

Through expanded Model Context Protocol (MCP) support, developers can now bring Zoom context and capabilities into leading AI assistants without requiring users to open the Zoom interface:

  • Microsoft 365 Copilot: Now supported, allowing users to access Zoom capabilities and meeting context within Copilot.
  • Existing support: Anthropic Claude, OpenAI Codex, and Meta Muse integration already available.
  • Google Gemini: Support coming soon.

MCP support has also expanded beyond Zoom Meetings to other products in the Zoom portfolio, including Zoom Rooms and Workspace Reservation. Additional MCP servers are planned for Zoom Phone, Scheduler and Mail, Calendar, and Marketplace Build.

Developers can also use new Search and Ask APIs to query Zoom's unified context layer from their own applications and MCP-compatible agents, allowing retrieval of relevant information across indexed sources.

Platform Studio: Unified Developer Portal

Platform Studio serves as a single entry point for developers building with Zoom's expanded platform. The portal integrates Zoom AI Services, Video SDK, and Co-Browse SDK in one location, providing developers with consolidated access to tools for building AI-powered communications experiences.

Platform Studio and the new Zoom AI Services APIs are available now, as are the Search and Ask APIs.