Runway adds ByteDance's Seed Audio to all paid plans
By Steven Van ·
Paid users can now generate up to 120 seconds of speech, sound design and music from a text prompt inside Runway.
Runway just closed one of the last remaining gaps in its AI creative suite. On June 29, 2026, the company confirmed that ByteDance's Seed Audio 1.0 is now available on Runway for all paid plans, letting creators generate up to 120 seconds of dynamic speech, sound design, and music from a simple text prompt. For a platform best known for state-of-the-art video generation, adding a native audio engine turns Runway into a genuinely end-to-end tool: you can now storyboard the picture and its soundtrack in the same workspace.
Every Paid Runway Plan Gets Seed Audio 1.0
The rollout is notably broad. Rather than gating Seed Audio 1.0 behind an enterprise or top-tier subscription, Runway has made it available across all paid plans. That means solo creators, small studios, and larger production teams all get access to the same underlying audio model from day one. Seed Audio 1.0 is the audio-generation model from ByteDance's Seed team, and it launched in late June 2026 as part of a wider rollout across AI platforms including fal, WaveSpeed, Runware, Higgsfield, and now Runway. Its arrival inside Runway brings the model to a much larger creative audience already generating video in the suite.
120 Seconds of Speech, Sound Design, and Music From Text
The headline capability is generating up to 120 seconds of audio from a single text prompt. Seed Audio 1.0 is not a narrow text-to-speech engine. From text alone it can produce natural speech, sound design and foley-style effects, music, and ambient sound, and it can mix those layers together in one pass. A single prompt can define who is speaking, what they say and how they feel, what is happening in the background, and which effects should appear. Because it runs from text, you do not need to source separate voice actors, SFX libraries, or ambience beds to rough out a scene's audio.
How Seed Audio Complements Runway's Video Generation
This is where the integration matters most. Runway is an AI creative suite for generating and editing video, images, and 3D, and it is best known for its advanced video models. Until now, audio was largely a reference input or a step handled in an external editor. With Seed Audio 1.0 running natively, creators can add a full audio layer directly alongside their generated footage. Voiceover, sound design, and ambient beds can be produced in the same tab as the video, keeping the workflow from prompt to audio-complete clip inside one platform. For anyone building short-form or cinematic AI video, that removes a round-trip to a separate audio tool for the bulk of a scene's sound.
What Seed Audio 1.0 Does Well and Where It Fits
Beyond the Runway integration, the broader Seed Audio ecosystem is built for expressive, controllable output. Capabilities include multi-speaker dialogue, voice cloning from reference audio, portrait-guided character voices, multilingual synthesis and dubbing, and fine-grained controls over delivery. That makes it a strong fit for narrated explainers, character dialogue, promos, and atmospheric scene sound. For creators who want a fully composed musical score, a dedicated music tool may still be the better choice, but for speech, effects, and ambience that sit under AI video, having Seed Audio 1.0 one click away in Runway is a meaningful upgrade to the workflow.
Sources: Runway announcement (X), AlphaSignal, fal (Seed Audio 1.0).