Freeway ML Feature Updates, November 11th

We released some dramatic improvements to image-to-image operations. The update improves Style changes, Variations, Prompt + Image, and Enhance operations to deliver a more consistent image pose and color palette. We've also launched our Discord bot, allowing linked users to generate images by typing /imagine.


Table of Contents

🖌 Improved Image Style and Variations

With our latest update, changing the style of an image is extremely fun; it’s one of our team’s favorite activities in the app today. Our improved style transformations can take a drawing and make it real or transform a photo into a different art form. This works for both images generated from Freeway ML or uploaded content. Hint, play with the creativity slider to get different results; for the images below, the creativity slider was low to ensure the style maintained a consistent form.

Style changes can become even more powerful when you combine them with variations. If you have an image you like, you can create a variant and then remix the variants with a new style to generate something completely unique!

Also - changing the image Style now generates 3 images instead of just one!

🌄 Prompt + Image Updates

We've also updated the application to improve consistency when editing an existing image with a text prompt. Now adding a prompt to the image can guide the engine to transform the output in a way that gives the creator directional control but maintains key image characteristics. This feature is extremely powerful, giving you the control to transform the result without compromising the pose or color temperature.

✨ Enhance Improvement

We've also given the Enhance operation a "facelift." With this update, Enhance delivers improved results on facial fixes, consistently removes blemishes, and reduces some of the grainy artifacts on clothing and in backgrounds. Overall, this operation should deliver more detail and help clean up aberrations.

✨ Introducing our Discord Bot

We've seen many community members try to call the AI image generator directly from Discord. This week we launched a bot where /imagine [prompt] posts generated images in our Discord server. We see this as the first step of many towards sharing image generations publicly.

✨ Community Feedback

We've received a ton of feedback in our channel about watermarks, bugs, and pricing. We take every comment to heart, and we are working to incorporate your feedback. This week we reduced the size and opacity of our watermark, iterated on several image quality updates, and squashed bugs from user feedback. Thanks again to everyone in the community that has expressed their thoughts!

⚡️ That's all for this week. Keep on creating!

AI Image Generator – create realistic images from text

Start creating for free

Keep reading

View all posts

Start Generating Magic ✨

Get started with Freeway ML's AI image creation tools

Start creating for free