New Midjourney Alternative but UNLIMITED and FREE! (Tutorial)

Matthew Berman
15 Feb 202308:19

TLDRMatt Berman introduces leonardo.ai, a free and unlimited alternative to Midjourney for creating AI-generated images. The platform offers a web interface and a variety of models, including photorealistic and paper art styles, allowing users to achieve high-quality and specific results. Berman demonstrates how to use the platform, highlighting its fine-grained control over image generation, the ability to remix images, and the creation of icon sets. He also discusses the platform's community models, where users can train their models with their images, and the prompt generation feature. Berman concludes by praising leonardo.ai as a superior and cost-free alternative to Midjourney, inviting viewers to share their creations and ask questions in the comments.

Takeaways

  • 🎨 **Leonardo.ai Introduction**: Leonardo.ai is a new platform for creating AI-generated images with more control and style variety compared to Midjourney, and it's free.
  • 🖥️ **Web Interface**: Unlike Midjourney, Leonardo.ai offers a web interface which is considered more convenient than using Discord.
  • 🌟 **Community Creations**: The platform showcases top creations by community members, displaying various styles and high-quality images.
  • 🔍 **Image Information**: Each image on Leonardo.ai provides detailed information including the model used for generation.
  • 🎭 **Fine-Grained Control**: Users have more precise control over image generation, allowing for closer results to the envisioned outcome compared to Midjourney.
  • 📈 **Featured Models**: Leonardo.ai features various models, including photorealistic and paper art styles, which users can select for their creations.
  • 🔄 **Remix Feature**: Users can remix any image directly on the platform, offering flexibility in creating variations of existing artwork.
  • ⏱️ **Rendering Time**: The time taken to generate an image on Leonardo.ai is comparable to Midjourney, and sometimes faster.
  • 🛠️ **Fine-Tune Models**: The platform includes fine-tune models that allow for more lifelike and stylized outputs, such as anime or isometric sci-fi buildings.
  • 🏗️ **Community Models**: Users can train their models using their own images, which can then be used immediately for image generation.
  • 📝 **Prompt Generation**: A unique feature of Leonardo.ai is the ability to generate ideas for base prompts, aiding users in the creative process.
  • 🖼️ **AI Canvas**: A new feature in beta allows users to manipulate canvas size and build upon images, offering more creative freedom.

Q & A

  • What is the main advantage of using Leonardo.ai over Midjourney for AI-generated images?

    -Leonardo.ai offers more fine-grained control over every aspect of image generation, allowing users to get closer to their envisioned result compared to Midjourney.

  • What is the first noticeable feature of Leonardo.ai that Matt Berman mentions in the video script?

    -The first noticeable feature is its web interface, which Matt Berman finds more convenient than using Discord.

  • How does Leonardo.ai allow users to interact with the community's creations?

    -Leonardo.ai allows users to click into any image, view information about it, see the model used, and even remix any image directly from the community's creations.

  • What are some of the different styles available in Leonardo.ai?

    -Leonardo.ai offers various styles including photo-realistic, paper art style, and isometric sci-fi building styles, among others.

  • How does Leonardo.ai enable users to create icon sets?

    -Leonardo.ai allows users to create icon sets without having to generate one icon per image and then combine them using external software like Photoshop. The platform generates ready-to-go icon sets.

  • What is the process for training a personal model in Leonardo.ai?

    -Users can upload their own images, which take about 30 minutes to train. Once trained, the model can be used immediately.

  • How does Leonardo.ai assist users in generating a base prompt for image creation?

    -Leonardo.ai has a prompt generation feature that provides users with ideas for their base prompt, starting with a basic concept and expanding on it.

  • What is the AI Canvas feature in Leonardo.ai?

    -AI Canvas is a feature that allows users to select the canvas size, add to certain images, remove certain images, and build off of them. It's a beta feature that enables users to erase and add to the image.

  • What does Matt Berman suggest about the image quality of the AI Canvas feature?

    -Matt Berman suggests that the images generated by the AI Canvas feature tend to be more odd and blurred between different features of the image, and not as high quality as the other features of Leonardo.ai.

  • How can users customize their feed on Leonardo.ai?

    -Users can customize their feed by going into their settings and specifying their interests, which affects what they see in the feed. They can also toggle NSFW content on and off.

  • What is the significance of the 'guidance scale' in the image generation process?

    -The guidance scale determines how strongly weighted the user's prompt is during the image generation process, influencing the level of detail and the adherence to the user's instructions.

  • How does the 'step count' affect the generated images?

    -The step count refers to the number of steps in the image generation process. A higher step count results in more detailed images, but it also takes longer to render.

Outlines

00:00

🎨 Introduction to Leonardo.ai for AI Image Generation

Matt Berman introduces Leonardo.ai, a platform for creating AI-generated images in various styles with greater control than mid-journey, and it's free. He discusses the convenience of its web interface over Discord and showcases examples of community creations. The platform allows users to remix images, control every aspect of image generation, and select from different models. Matt demonstrates creating a paper art style image of a Japanese village and compares the process to mid-journey, highlighting the specificity and fine-tuning capabilities of Leonardo.ai.

05:00

🚀 Advanced Features of Leonardo.ai: Models, Training, and Prompts

The video continues with a demonstration of AI image generation in Leonardo.ai, including selecting the number of images, dimensions, guidance scale, step count, and tiling. Matt also shows how to use a base seed for an image. He then explores community models, the ability to train a model with personal images, and the immediate usability of these models. The output quality is shown to be dependent on the resolution of the training images. Additionally, Matt introduces prompt generation, a feature not available in mid-journey, which assists users in creating base prompts. He also discusses the AI canvas feature, which allows for the manipulation of canvas size and image composition, though noting it's in beta and doesn't produce as high quality as other features. The video concludes with Matt sharing some of his favorite creations from Leonardo.ai and encouraging viewers to ask questions, share their creations, and subscribe for more content.

Mindmap

Keywords

AI-generated images

AI-generated images refer to visual content created using artificial intelligence algorithms. In the context of the video, it involves using a platform like Leonardo.ai to generate images in various styles based on user prompts. The video emphasizes the high quality and control that AI provides over the image generation process, which is central to the video's theme of exploring an alternative to Midjourney.

Leonardo.ai

Leonardo.ai is an online platform mentioned in the video that allows users to generate images using AI. It is presented as a more convenient alternative to using Discord for similar purposes. The platform is highlighted for its web interface and the ability to fine-tune the image generation process, which is a key focus of the video tutorial.

Web interface

A web interface is a way for users to interact with a software application via the internet, using a web browser. In the video, the presenter finds Leonardo.ai's web interface more convenient than using Discord, which implies a more user-friendly and accessible approach to generating AI images. It is a significant aspect of the platform's appeal as discussed in the video.

Control over prompts

Control over prompts refers to the ability of a user to influence the output of an AI image generation system by providing specific instructions or 'prompts'. The video emphasizes that Leonardo.ai offers more fine-grained control, allowing users to get closer to their envisioned image compared to other services like Midjourney.

Featured models

Featured models in the context of the video are the different styles or presets available within Leonardo.ai for generating images. The platform offers a variety of models, such as 'Deliberate 1.1' for photorealism and 'paper art style', which are used to create images in distinct visual styles. These models are a core part of the video's demonstration.

Remix

To remix, in the context of the video, means to take an existing AI-generated image and modify it to create something new. The presenter demonstrates this by changing an RPG icon set to an FPS game, showcasing the flexibility of Leonardo.ai to adapt and create new content based on user preferences.

Stable Diffusion 1.5

Stable Diffusion 1.5 is a specific model or version of the AI algorithm used within Leonardo.ai for generating images. The video mentions it as an option for users to select when generating their images, indicating the variety of tools available within the platform for creating different styles of images.

Negative prompt

A negative prompt is a user instruction given to an AI image generation system to exclude certain elements from the generated image. In the video, the presenter uses a negative prompt to specify that they do not want any faces shown in the generated images, demonstrating the level of control users have over the output.

Icon sets

Icon sets refer to a collection of images that are related in style and theme, often used for graphical user interfaces or design elements. The video highlights Leonardo.ai's ability to generate icon sets efficiently, without the need for additional software like Photoshop, which streamlines the process for users.

Fine-tune models

Fine-tune models are AI models that have been adjusted or customized to perform better on a specific task or to meet certain user requirements. In the video, the presenter discusses using platform and community models within Leonardo.ai, which can be fine-tuned to generate images that match the user's vision more closely.

Prompt generation

Prompt generation is a feature that assists users in creating the initial instructions or 'prompts' for the AI image generation process. The video describes how Leonardo.ai can generate ideas for base prompts, aiding users who may struggle with crafting effective prompts to achieve their desired image outcomes.

AI Canvas

AI Canvas is a feature within Leonardo.ai that allows users to manipulate and build upon AI-generated images, including resizing the canvas and adding or removing elements. The video presents this as a new and innovative tool for users to further customize their images, although it notes that the feature is still in beta and may not generate as high quality images as other options.

Highlights

Leonardo.ai is an AI image generation platform offering more control and style variety than Midjourney, and it's free and unlimited.

The platform has a user-friendly web interface, which is more convenient than using Discord.

High-quality, diverse styles of AI-generated images are showcased, including hyper-realistic and paper art styles.

Users can click on any image to view detailed information, including the model used for generation.

Leonardo.ai allows for fine-grained control over image generation, enabling users to achieve closer results to their vision compared to Midjourney.

Featured models on the platform include photorealistic and paper art styles, offering a wide range of creative possibilities.

The platform enables users to remix any image directly from the gallery.

Image generation with Leonardo.ai is as fast as Midjourney, and sometimes quicker.

Leonardo.ai stands out for its community-driven model training, allowing users to create their own models with a 30-minute training time.

The platform provides a unique feature for creating icon sets without the need for additional image editing software.

Users can fine-tune platform models to generate specific styles, such as lifelike anime or isometric sci-fi buildings.

Community models on Leonardo.ai are trained by users, offering a personalized and diverse set of styles.

The platform includes a prompt generation feature to help users come up with creative ideas for their base prompts.

AI Canvas is a new feature that allows users to generate, modify, and build upon images in a unique way.

Leonardo.ai provides settings to customize the user's feed based on their interests and allows toggling of NSFW content.

Matt Berman, the presenter, shares his favorite creations from Leonardo Ai, emphasizing the platform's capabilities.

The video encourages viewers to ask questions, share their creations, and engage with the content.

Leonardo.ai is positioned as a superior and cost-free alternative to Midjourney for AI image generation.