Adobe has officially launched the Firefly audio generation tool, which includes "Generate Music," "Generate Voice," and "Generate Sound Effects." Users can directly create commercially secure audio content within the Firefly desktop and mobile web application, further integrating AI audio capabilities into the video creation workflow.

A recent survey conducted by Adobe and the Berklee College of Music found that 100% of the surveyed video creators, musicians, and marketers use music in their videos, with 32.7% already using AI-generated music; 43.2% cited legal and copyright risks, and 38.9% considered licensing fees as major obstacles.

QQ20260821-093447.jpg

To address these pain points, Firefly's "Generate Music" can generate original and fully licensed music based on video content, atmosphere, rhythm, and duration, offering four options for users to edit and choose from. Adobe stated that generated content can be used in commercial works without additional subscriptions.

"Generate Voice" can convert scripts into natural narration, supporting Adobe Firefly voice models and ElevenLabs models, and allows users to adjust voice effects through tags such as expressiveness and pronunciation; "Generate Sound Effects" supports text prompts or recorded sounds as references, generating sound effects that match the visual action, rhythm, and atmosphere.

At the same time, Adobe continues to expand the Firefly third-party model ecosystem, adding Gemini Omni Flash, which can combine video, audio, and images for content creation. Previously, Firefly had integrated Runway Aleph2.0 and Kling2.0, and now the model lineup includes Google, Kling AI, Luma AI, OpenAI, Runway, and Adobe's own models. Adobe is further perfecting Firefly's one-stop AI creation workflow by integrating multiple models and multimodal audio-visual capabilities.