ElevenLabs drops Image & Video beta with Sora and Kling integration

Making an AI video has always been a painful game of digital hopscotch. You generate an image on one site, bring it to a video generator on another, head to ElevenLabs for the voiceover, and finally drag everything into an editor to manually line up the lip movements.

ElevenLabs just launched ElevenLabs Image & Video in beta. The platform, best known for its realistic voice models, is now bundling top-tier visual and audio generators into a single, unified creative workspace.

Leading visual models under one roof

Instead of making users choose between platforms, ElevenLabs pulled in the heavy hitters from across the industry.

For still images, the update integrates models like Nanobanana, Flux Kontext, GPT Image, and Seedream. Creators can use these for visual concepting, storyboarding, or source material for video generation.

For video generation, it brings in Veo, Sora, Kling, Wan, and Seedance. Users can generate clips across these models, upscale visual assets, and string together multiple takes into a coherent story.

Crucially, it also features automated lip-sync. You can take any generated video and match the speaker mouth movements directly to ElevenLabs synthetic voices without bouncing out to separate tools.

Finishing the cut in Studio

Once visual assets are ready, everything feeds directly into ElevenLabs Studio for post-production.

Basically, creators no longer need to switch over to an external timeline editor just to add audio layers.

Inside Studio, users can add voiceovers from the standard library or custom voice clones, generate custom background music, and layer in sound effects. Everything lives on a single editing timeline where timing and narration can be dialed in before final export.

Why it matters

For solo creators, marketers, and video agencies, this eliminates huge chunks of context switching and subscription bloat. If your workflow involves talking-head characters or narrated social clips, having high-end generation, lip-sync, and audio under one roof is worth trying right now.

We think this is an exceptionally clever tactical pivot. Instead of burning hundreds of millions to train a proprietary foundation video model from scratch to challenge the giants, ElevenLabs simply embraced them all. It turned rivals into backend utilities while defending its own moat in audio and workflow.

It also highlights where generative video competition is heading. Generating pretty clips is becoming a commodity; what creators actually need is a finished file ready for export. While managing multi-vendor API costs and stylistic consistency across different engines might get tricky, ElevenLabs just moved one giant step closer to becoming the definitive operating system for multimodal storytelling.