Images and media
Distinguish image input, image generation, video generation and design exports.
Four different uses
| Use | Entry |
|---|---|
| Understand an image | Attach it to a conversation with a vision-capable model |
| Generate or edit images | Use an enabled image-generation capability and a supported model |
| Generate video | Use a supported model/tool and follow its task state |
| Export a design | Render a Design project to an image, PDF or video |
Design video export is not the same as a video-generation model call. Availability depends on the model, enabled tools, account permissions and service status.
Image input
Add the image, confirm that its attachment appears, and explain what to inspect. Follow the actual image-sharing consent settings. If image understanding is disabled or unsupported, select a capable model or adjust the relevant setting. A text reply does not prove that the image was received.
Generate and inspect results
Specify content, dimensions or resolution, duration and intended use. Review model and cost information before execution. Track the task to completion, then inspect the actual file. An accepted request is not a completed artifact.
Images are commonly priced per image, and video per second and resolution; use the selected model's current rules. Text-token prices do not price media generation.
Troubleshoot
- Missing tool: check media settings, model capabilities and account access.
- Missing attachment: check upload state and the sent message.
- Missing download: verify the task's terminal state and artifact location.
- Failed Design export: inspect the export record and renderer; see Design.