Keeping multimodal experimentation in one place

Teams building with image generation models or vision pipelines often hit the same problem. The model produces an image, but you cannot see it where the prompt lives.

That makes reviewing quality manual, comparing runs messy, and iteration slower than it should be.

We just shipped native image rendering inside Datasets and Prompt Workbench. Generated images now appear directly next to the prompts that created them.

This means:

• Faster output review
• Easy visual comparison across runs
• Iteration without switching tools or losing context

Prompting, generating, reviewing, and experimenting now happen in one place. Multimodal workflows finally get tooling that matches how they actually work.

Mutimodal - Image Generation in Datasets & Prompt: https://shorturl.at/athOG