The platform analyzes videos to generate text based on their content using a multimodal approach. This method analyzes the visuals, sounds, spoken words, and relationships between them. For guidance on writing prompts, see the Prompt engineering page.
The page uses a two-panel layout. Select a video, enter a prompt, and configure parameters in the left panel. Results appear in the right panel.
When you open the Analyze page, you can start in one of the following ways:
In the left panel, select the video preview area to choose a video that you’ve already uploaded, or upload a new video directly from the dialog.
In the prompt field, enter a prompt that describes what you want to analyze. Your prompts can be instructive, descriptive, or phrased as questions. To include an image in your prompt, select the image icon at the bottom of the prompt field. In the Add an image dialog, upload a new image, choose an existing image asset, or paste an image link.
(Optional) Adjust the parameters under Advanced Settings. To reset all parameters to their default values, select the Reset button.
(Optional) To receive a structured JSON response, use the structured_output section under Advanced Settings. You can enable structured output in two ways:
To customize the schema, select the Edit in Builder button. The JSON Schema Builder modal opens with two panels. Define your schema in the Builder panel (visual interface) or the Editor panel (raw JSON). Changes in either panel are reflected immediately in the other.

Select the Analyze button. The platform displays progress messages while processing. You can leave the page and return later to see the results.
Results appear in the right panel. Select the Visual or JSON tab to switch between views.
The Visual tab shows the generated text.

The JSON tab shows the raw JSON response.

If the generated text includes timestamps, select one in the Visual tab to play that moment.

Select the View Code button to see the request as cURL, Python, or JavaScript. Copy and paste the snippet into your application.

To analyze the same video with a different prompt, edit the text in the prompt field and select the Analyze button to rerun the analysis.