ElevenLabs MCP Adds Voice, Music, Image, and Video Generation Inside Chat Assistants

ElevenLabs expanded its Model Context Protocol (MCP) server on September 14, 2026, moving it beyond agent management into direct creative generation from inside chat assistants like Claude, ChatGPT, and Cursor.

Details

• From management to creation: ElevenLabs framed the update directly — the MCP “already let you manage your agents from the assistants you work in every day. And now, it does more than that - it lets you create”
• Voice and speech tools: text-to-speech voiceover generation in any voice from the library, plus speech-to-text transcription with speaker labels and timestamps across 99 languages, and dubbing that preserves original speaker characteristics
• New media types: the MCP now also generates music, sound effects, images, and lipsynced video
• Broad client support: the MCP is live across Claude, ChatGPT, Cursor, Grok Bot, Hermes, and other MCP-compatible assistants, installed via each platform’s connector directory using OAuth — no API keys or manual server setup required
• Model breadth: the connector integrates more than 50 underlying models
• Workspace integration: everything generated through the MCP lands in the ElevenCreative workspace for further editing in Studio, with workspace admins controlling tool access and data residency settings

What happened next

The update turns ElevenLabs’ MCP server into a general-purpose creative backend for any assistant that supports the protocol, rather than a narrow agent-ops tool. It arrived three days after the Music v2.5 model launch and two days before the Reception product announcement, part of a dense mid-September stretch of ElevenLabs product releases spanning models, integrations, and new applications.