A serious creative upgrade just landed

This update is smaller in count, but much bigger in impact.

GPT-image-2 is now live in Youanai, Lisa has new controls for model and reasoning depth, Lisa has better visibility into your workspace, and video generation should be more reliable where it matters.

GPT-image-2 is now the default image model in Youanai

We added GPT-image-2 to Youanai the day it became available and made it the default image model across the product.

This is not a tiny quality bump. For a lot of business creative, it feels like the difference between an interesting AI image and an image you can actually use in real marketing.

Some customers who have already tried it have told us the jump feels insane. Better composition, stronger creative direction, cleaner outputs, and far less of the old AI-image weirdness.

Choose Lisa's model and reasoning level

Lisa now has model and reasoning controls, so you can choose both the model she uses and how deeply she should think before she starts working.

Use lighter settings for quick edits, small rewrites, and simple tasks. Use deeper reasoning when you want Lisa to think through strategy, campaign direction, content planning, or more complex work.

OpenAI models are available now, and we're already building support for Anthropic and Google models next.

We also improved how Lisa sees your workspace and posts, so she can ground more of her help in what already exists inside your account instead of treating every request like it starts from zero.

Lisa is easier to start from Home

We refreshed the Home experience so Lisa is more central and easier to reach when you first open Youanai.

You can start a Lisa request directly from Home, attach references, and set creative preferences without hunting around for the right place to begin.

The goal is simple: when you know what you want, Lisa should be right there and ready to help turn it into work.

Video generation should be more reliable

We tightened up video generation, especially around UGC-style Sora videos, portrait formats, and reference-image handling.

That means fewer jobs should fail for preventable sizing reasons, and vertical UGC-style videos should behave more predictably.

It is one of those updates that may not look flashy from the outside, but it makes the creative workflow feel much less fragile.

The theme this round is practical creative leverage: better images, smarter Lisa, easier starting points, and fewer broken edges in video generation.

If you try GPT-image-2 in Youanai, hit reply and tell us what changed for your business. We're especially curious where the new image quality suddenly makes a workflow feel possible.

Keep building, The Youanai Team