Leonardo AI Just Launched Two Game-Changing Updates: Video & Custom Models

midjourney blog post image

Update 1: Motion 2.0 for AI Video

Generating video from AI is now much easier with Motion 2.0. This update is built right into the AI creation interface. You can find the “Video” option nestled between “Image” and “Flow State” at the top of the screen.

How to Create AI Videos

There are a couple of ways to generate video with Motion 2.0:

  • Prompt Only: You can simply type in a prompt and Leonardo AI will generate a short video based on your text.
  • Image to Video: Select an image you’ve already created in Leonardo AI, open it, and click the “Image to Video” option. This uses your image as a starting point.

Prompting Tips for Video

Unlike AI images where detailed prompts work well, shorter, more precise prompts are better for AI video. Think about the main action or scene you want to see.

Exploring Motion 2.0 Examples

Using the image-to-video feature with an image like a superhero holding a gun can produce dynamic results, like a handheld camera panning right. Experimenting with prompts alone can also yield interesting mini-videos, such as a snowboarder going down a mountain or an FPV drone flythrough.

Comparing videos generated with just a prompt versus those using a starting image shows the different possibilities. Both methods can create impressive short clips, adding movement and life to your AI creations.

Update 2: Training Custom Models (LoRAs)

Leonardo AI now lets you train your own custom models. This feature is similar to LoRAs in other AI art tools. You can train the AI on specific styles, objects, and soon, characters.

Finding the Training Feature

The new training option is on the left-hand side menu, under “Advanced Models and Training.” Click “Train New Model” to get started.

How to Train Your Own Model

Training requires a dataset of images. Here’s the basic process:

  • Gather Images: You need multiple images of the style or object you want to train. Using Leonardo AI itself (like with Flow State) is a great way to generate these reference images. Generate several similar but varied images of your target subject (e.g., a specific type of purse).
  • Upload Dataset: Name your model and upload your collection of images. You can upload up to 50 images.
  • Set Details: Choose your base model (like FluxDev) and specify the category (Object, Style, Character). You might need a trigger word depending on the base model.
  • Start Training: Begin the training process. This can take some time, like around 30 minutes.

Using Your Trained Model

Once trained, your custom model appears in your list of models. When creating a new image with your trained model selected, a trigger word or tag (often the name you gave it) will appear in your prompt box. You can adjust its strength to control how much influence it has on the final image.

Custom Model Examples

Training a model on an object, like a black designer purse, can show impressive results. When you use the trained model with prompts like “beautiful woman walking… holding a purse,” the generated images consistently feature the trained object. You can also test the model with different prompt styles (product shots, fashion models) and model settings (Dynamic, Fashion, Portrait, Cinematic) to see its versatility.

The AI does a great job of incorporating the trained object, even if the source images weren’t perfectly identical (e.g., some had keychains, some didn’t). It adapts the core features of the object across various scenes and styles.

Conclusion

These two updates, Motion 2.0 for AI video and custom model training, bring powerful new capabilities to Leonardo AI. While these features are still being improved, they offer creators exciting ways to explore new forms of AI art. Keep experimenting, and get ready for even more updates in the future.

🖥

Do This Locally with Logen

Skip the subscription. Generate images like these on your own GPU — private, unlimited, no cloud dependency.

Try Logen Free →

Midjourney vs Logen

FeatureMidjourneyLogen
Cost$10-60/moFree & Pro plans
PrivacyCloud (images stored)100% local
GenerationsLimited by planUnlimited
Internet RequiredAlwaysNo
Custom ModelsNoYes (any ComfyUI model)
GPU RequiredNo (cloud)Yes (local)
Ease of UseDiscord botDesktop app
Try Logen Free →
FREE 11-PAGE PDF HANDBOOK + LOCAL AI FIELD NOTES

Get the Free 11-Page Local AI Image Generation Handbook

See what GPU VRAM you need for FLUX.1-schnell and SD 1.5, how checkpoints and LoRAs work, and five repeatable workflows on Windows and Linux x64. Enter your email to open the 11-page PDF right away and get our six-part weekly local AI field notes.

6–24 GB VRAM Tables Cloud vs. Local Math 5 Studio Workflows
How signup uses your email

We register your email with the TitanXT newsletter on beehiiv to send the 11-page PDF link, our six-part weekly local AI sequence, and occasional Logen updates. You can unsubscribe in one click from any email or by writing to support@titanxt.io. Read beehiiv's privacy policy (new tab).

Scroll to Top