DaVinci AI

A workspace for generated images and video

DaVinci AI is HubX's creative platform for generating and editing visual media. Users choose an image or video model, describe the result they want and refine the creation within the same service. Web and mobile entry points make it relevant to content creators, designers and small marketing teams preparing visuals for different channels.

This is a separate product from Blackmagic Design's DaVinci Resolve video editor. DaVinci AI brings third-party generation models into its own interface and subscription arrangement. Access to a model through this service should not be confused with a subscription to the model provider's separate application.

Start with text or an image reference

The image-generation guide describes selecting a model, writing a prompt and optionally uploading an image for image-to-image generation. Subject, colour, lighting and style details help define the brief. Once a result is generated, users can download it or continue editing and revising the prompt.

Model choice matters because capabilities, output style and generation costs differ. For a useful comparison, hold the brief steady while changing one setting or model. A reference is guidance for a new image rather than an assurance that a product label, face or composition will remain exact.

Refine the scene and adapt its format

DaVinci's editing documentation lists upscaling, background changes, resizing, restyling, relighting and replacing or removing objects. These tools address different stages of preparation: removing a distraction changes the scene, while resizing adapts an asset for a different placement.

The video workflow accepts a text brief or an image to animate. Users select a video model, generate, preview and download or revise the result. The guide notes that video typically consumes more credits than image generation, with usage and processing time depending on the model, resolution and length.

A small campaign example

Suppose a local maker needs a visual direction for a new candle collection. Prepare an original product photograph and describe a simple scene with a neutral surface, soft side lighting and space for a headline. Generate a small initial selection rather than requesting every campaign format immediately.

Compare the candle's shape and label with the photograph. If a scene works, remove a distracting object or adjust the background before adapting the crop for each placement. Use the approved image as the starting point for a short video experiment, then inspect movement, reflections and text throughout the clip. Keep any final promotional claims in a separately reviewed layout instead of relying on generated lettering to communicate product facts.

Free requests and paid credit access

The access guide describes three initial generation requests without payment, followed by paid access for continued creation and broader model use. The website explains that tools consume different amounts of credit, subscription credits reset with the billing cycle, and additional credits require an active subscription. Check the current account offer rather than treating an advertised maximum image count as equivalent to every model or video task.

What should creators review?

Inspect the actual export for altered identities, spelling and invented product details. DaVinci's homepage lists audio generation as forthcoming, so audio is not included here as an established feature. Review the selected model's requirements, account terms and permissions for uploaded material before using a result in client work. Generated examples are creative possibilities, not evidence of guaranteed output quality.