Senior Software Engineer

Sonilo and fal have launched Sound Effects 1.0, a model that generates realistic sound effects from video or text. The system analyses on-screen motion, timing and scene context to create synchronised audio tracks matched to footage, supporting videos up to three minutes long. With video input, the model produces a finished audio track aligned to what appears on screen. Text input allows developers to describe and generate specific sound effects directly. The system aims to address gaps in AI video production, where footage often requires significant manual audio work. fal will serve as the exclusive API launch partner, providing developers with immediate access through its infrastructure. Sound Effects 1.0 expands the existing Sonilo-fal partnership, which already includes Sonilo Music v1.1 for video-to-music generation. The model is designed for short-form content, advertisements, gaming footage and narrative scenes. Sonilo, based in San Francisco, is backed by B Capital.

Sonilo has launched its video-to-music AI model on fal.ai, enabling creators to generate commercially licensed soundtracks from video footage in seconds. The model analyses a video's pacing, motion and emotional arc to automatically compose original music matching the exact duration. The San Francisco-based startup reports that in internal tests, editors accepted the model's first attempt 87% of the time, whilst videos scored with Sonilo showed a 16% increase in engagement. The platform can score videos up to 600 seconds long and generates multiple soundtrack options for each clip. Trained on professionally licensed content including Shutterstock's music catalogue, Sonilo's outputs are available for commercial use. The company is backed by B Capital and previously integrated with ComfyUI in April. A text-to-music model is also available on fal.ai.

Sonilo brings licensed AI music generation to fal.ai for automated video soundtracking. Published on: Jun 22, 2026 As AI-generated video becomes increasingly mainstream, audio remains one of the most challenging elements of content production. While creators can now generate visuals, edit footage, and create voiceovers using artificial intelligence, finding commercially safe music that matches a video's tone, pacing, and duration often remains a manual process. Sonilo is aiming to solve that problem with the launch of its video-to-music model on fal.ai, allowing developers and creators to generate licensed soundtracks automatically from video content. AI music startup Sonilo has expanded its reach into the growing generative media ecosystem through a new integration with fal.ai, making its video-to-music and text-to-music models available to developers, content platforms, and creative technology providers. The launch positions Sonilo within a rapidly evolving segment of the creator economy where artificial intelligence is increasingly automating complex production workflows. While AI tools have transformed image generation, video creation, and content editing, music licensing and soundtrack production remain relatively fragmented processes that often require creators to navigate stock libraries, licensing agreements, and manual editing tasks. Sonilo's technology addresses this challenge by analyzing video footage directly and generating original music designed to match the visual content. Instead of relying on text prompts, the platform evaluates factors such as pacing, motion, scene transitions, and emotional tone before composing a soundtrack synchronized to the video's duration. The approach reflects a broader trend across generative AI platforms: reducing the number of manual steps required to create publish-ready content. For content creators, marketing teams, social media publishers, and video production platforms, music selection can often become a bottleneck in the production process. A soundtrack that is too long, too short, or emotionally mismatched can reduce audience engagement and require additional editing time. Sonilo's system attempts to eliminate those friction points by generating music tracks that align with the exact length of a video. The resulting soundtrack is delivered as a separate audio layer, allowing editors to adjust volume independently while preserving dialogue, narration, interviews, and sound effects already present in the source footage. One of the more notable aspects of the launch is its focus on licensing and commercial usage rights. Copyright concerns continue to be one of the most significant challenges facing the generative AI industry, particularly in creative sectors involving music, video, and intellectual property. Sonilo says its models are trained on professionally licensed content, including music assets from Shutterstock, with participating musicians compensated for their contributions. That licensing foundation may prove increasingly important as brands and enterprises adopt AI-generated creative assets at scale. Many organizations remain cautious about deploying AI-generated content without clear commercial rights protections, particularly when content is intended for advertising campaigns, branded media, or monetized digital channels. The integration with fal.ai expands Sonilo's accessibility to a wider ecosystem of AI developers. fal.ai has emerged as a growing infrastructure layer for generative media applications, providing APIs and deployment tools that allow developers to integrate AI models directly into products and workflows. Through the platform, Sonilo's video-to-music model can generate soundtracks for videos up to 600 seconds in length. The company has also made its text-to-music model available, offering creators prompt-based generation capabilities alongside advanced controls that support multiple moods, genres, and structural changes across different sections of a composition. The launch arrives at a time when multimodal AI systems are becoming a major focus across the technology sector. Companies including Google, Microsoft, Adobe, and Amazon are investing heavily in tools capable of combining text, image, audio, and video generation into unified workflows. For creative technology vendors, the opportunity extends beyond content creation. Enterprises increasingly want AI systems that can automate entire production pipelines rather than individual tasks. Music generation tied directly to video content represents one example of how AI models are evolving from standalone tools into integrated production infrastructure. According to Sonilo, internal testing found that editors accepted the first generated soundtrack on 87% of evaluated clips. The company also reported a 16% increase in engagement metrics for videos scored using its technology, suggesting that soundtrack quality remains an influential factor in audience retention and content performance. While such results will likely require validation across broader production environments, they highlight an important trend: AI-powered optimization is moving beyond visuals and into audio experiences that can influence viewer behavior. The launch also follows Sonilo's earlier integration with ComfyUI, signaling a strategy focused on becoming a foundational music generation layer for AI-native creative ecosystems. As generative video adoption accelerates across marketing, advertising, entertainment, and social media sectors, automated soundtrack generation may become a critical component of next-generation content workflows. For developers building AI video platforms, creator tools, editing software, and multimodal content systems, Sonilo's arrival on fal.ai offers another example of how specialized AI models are being assembled into increasingly sophisticated media production stacks. Market landscape. The AI-generated media market is expanding rapidly as organizations seek to automate content production workflows. According to Gartner, generative AI continues to be among the fastest-growing enterprise technology categories, while IDC projects significant investment in AI-powered content creation platforms over the next several years. Within the creator economy, audio generation remains one of the least automated production stages compared with image and video generation. As multimodal AI adoption grows, technologies capable of synchronizing music, voice, visuals, and editing workflows are expected to become key components of enterprise content operations, digital marketing platforms, and creator-focused SaaS ecosystems. Top insights. * Sonilo has launched its licensed AI video-to-music model on fal.ai, enabling automated soundtrack generation for video creators, developers, and AI-powered media platforms. * The platform analyzes pacing, motion, and emotional context within videos to generate original music synchronized to exact video durations. * Commercial licensing remains a key differentiator, with Sonilo training models on licensed music catalogs and offering commercially usable outputs. * The integration strengthens fal.ai's growing ecosystem of multimodal AI tools supporting next-generation video production and creative automation workflows. * Demand for AI-powered media infrastructure continues rising as enterprises seek faster, scalable methods for producing video, audio, and marketing content.