Magic Hour is a web-based AI studio that combines video generation, image creation, and audio tools in one platform, aimed at creators, marketing teams, and developers who want to produce media without juggling separate apps. The platform offers over 100 tools, from text-to-video and face swap to voice generation and image editing, and it provides a REST API with SDKs for Python, Node, Go, and Rust for programmatic workflows.
Magic Hour organizes its tools around three media types: video, image, and audio. On the video side, you can generate clips from text or animate still images, swap faces, sync lips to audio, and turn photos into talking avatars. Image tools cover generation from prompts, background removal, upscaling, and generative fill. Audio tools include an AI voice generator, voice cloner, and music generator. The idea is to keep assets in one place, so you can start with a prompt or an upload and then refine the result with related tools without exporting to another application.
The studio includes more than 25,000 templates sized for different channels, which can help teams move from a blank page to a publishable asset more quickly. For teams and agencies, Magic Hour supports campaign workflows where a single reusable process can personalize content at scale, such as face swap or virtual try-on for product visuals. The platform also provides starter recipes for product videos and lip-sync projects, giving users a defined path rather than a loose collection of tools.
Developers can access many of the same tools through a REST API, with official SDKs for Python, Node, Go, and Rust. The API supports text-to-video, image-to-video, face swap, lip sync, and image generation, among other endpoints. A free sample request is available before creating an account, and the documentation includes an evaluation guide for comparing cost per accepted output. API usage draws from the same credit balance as the web app, and there are no separate rate or concurrency limits for API calls, though web app capacity depends on the plan.
Core tools are free to try without signing up. Text-to-video and image-to-video share three free generations per day, each producing a 3-second clip with audio. Free video exports are 480p and include a watermark. Paid plans and credit packs remove the watermark and unlock higher resolutions, faster queues, larger uploads, and more concurrent generations. Unused credits never expire, and credit packs can be purchased without a subscription. The platform uses a single credit balance across web and API use.
Magic Hour states that it does not sell customer prompts, uploads, or outputs, and does not use them to train AI models. This no-training policy applies to free and paid accounts, including API and MCP use, without requiring an enterprise plan or opt-out. Personnel do not routinely review customer content; access is limited to specific cases such as investigating reported problems or complying with law. Users retain ownership of their inputs and, under the terms, own the outputs they create. Account deletion is available, and project deletion can be managed through the service.
Magic Hour suits creators who want a broad set of AI media tools in one interface, teams that need to produce campaign assets at volume, and developers building automated image and video workflows. The combination of a free tier, template library, and API access makes it a practical option for those who want to test capabilities before committing to a paid plan. The platform reports 4.8 million creators and 2,000 teams, including the Dallas Mavericks creative team, which used it to produce an anime-inspired hype video.