Astrea

Astrea is a video ad maker organised around projects. You start with a brief, an image or a video reference, and its Video Agent drafts the next ad while keeping the parts of the project worth reusing, such as the product story and the call to action. The homepage walks through the flow with a demo product, a water bottle. You describe the product, audience, offer and direction (its example is a 15-second vertical ad for commuters), review the script and production plan, then generate a candidate and refine it. The workspace has Chat, Canvas and Timeline views, so the brief, references and earlier decisions stay together. The creation link goes to a Google sign-in. The walkthrough is labelled illustrative, and generation quality and how faithfully a product is rendered aren't shown on the public pages.

This may suit a small marketing team exploring visual directions before making a finished ad, or one that needs fresh creative when an existing ad starts wearing out. The site frames it around that problem: try a new hook, offer or angle without rebuilding the project. A reference asset gives the concept a specific product to work around. The visible examples do not establish campaign performance or whether a generated result preserves factual product details.

Basic costs $19.90 a month for 1,000 credits and Pro $59.90 a month for 3,300 credits. Both include 720p and 1080p generation, video from a reference image, a priority rendering queue and clip downloads, and annual billing is advertised at up to 50 percent off. One-time credit packs start at $9.90 for 400 credits and stay valid for 365 days.

The privacy policy names the services behind the product. Sign-in uses Google, billing runs through Stripe, uploaded media is stored in Cloudflare R2, image generation goes through a 12AI/OpenAI-compatible API, and video generation runs on Alibaba Cloud DashScope. In practice, the reference images, videos and prompts you add are sent to those providers to produce results. Check the terms for who owns generated assets before using them in paid campaigns.

The project-based approach is the clearest idea here. Each new ad starts with the context of the last one instead of a blank prompt. Whether that saves time depends on output quality, which only a trial on your own product will show.