Start with a brief, a project or a set of sources. Follow the steps to a useful deliverable.
Choose a workspace with the media tools you need
From a brief to creative assets
Start with: A launch brief, brand colors and a reference image.
- Develop the concept and choose an image-generation service.
- Inspect the result and ask for a specific edit.
- Add another supported medium if needed, then save the files you will use.
Finish with: Creative files you can inspect and use in your project.
Generation services have their own capabilities and costs. Review every output before delivery.
Explore creative productionDesktop
Describe, adjust and export a 3D prop
Start with: A low-poly prop description, dimensions and adjustable parameters.
- Ask for a parametric model and inspect its live preview.
- Change a parameter with the controls and inspect the geometry from multiple angles.
- Export the GLB and open it in your next tool to check the result.
Finish with: An exported 3D asset and editable model source.
Keep the editable source for later changes, and use the GLB in your next 3D tool.
Explore Desktop 3D modelingWeb workspace
Research to a maintained client report
Start with: A client question, supporting sources and a dated data set.
- Gather the sources in a notebook and ask for a cited interactive report.
- Review the analysis, publish the report and deliberately choose its audience.
- When inputs change, review a revised report and manually republish to the same link.
Finish with: A client deliverable with reviewed evidence and a link you can keep updating.
Choose who can open the report before sharing. Republish when you want your client to see an update.
Explore the team workflowDesktop
Choose where inference runs
Start with: A project and a task suitable for your local or hosted model.
- Choose your local model server and complete a task.
- Choose a hosted model for the next task when it suits the work.
- Check which model ran, what project context it used and the resulting output.
Finish with: Useful work from either inference route in the same Desktop workspace.
Model and tool support vary. Some engines require a new conversation. External services still run externally.
Explore inference choice