Flows Agent builds multimodal pipelines in ElevenCreative

ElevenLabs integrates Flows Agent into ElevenCreative to build and edit multimodal pipelines connecting over 50 image, video, and audio models via text.

A conversational agent is integrated into the core of ElevenCreative Flows, ElevenLabs' nodal canvas that connects over 50 image and video models with proprietary audio components, including voice, music, SFX, and lip-sync. With Flows Agent, a natural language description is all it takes: the agent selects the appropriate models, places the nodes on the canvas, wires the connections between them, and executes the generations.

Iteration is then driven by conversation. Requesting a warmer voice, replacing a background, or adapting the production into Spanish only modifies the relevant segments of the pipeline, which is relaunched without full reconstruction. Cost control remains with the user: in assist mode, the agent requests validation before any costly operation. It can also run in the background, with the tab closed, and present a completed workflow upon return. Regaining control to manually adjust the nodes remains possible at any time.

ElevenLabs positions the tool for creators who chain modalities into complete pipelines—image, video, voice, music, and effects all in one place—and for marketing teams that multiply variants by product, language, or format. Flows Agent is available now in ElevenCreative, with an ElevenLabs account in hand.