Stability AI releases most powerful image generation models to date

Date:

Stability AI has announced the release of Stable Diffusion 3.5, marking a leap forward in open-source AI image generation models.

The latest models from Stability AI include multiple variants designed to cater to different user needs, from hobbyists to enterprise-level applications.

The announcement follows June’s Stable Diffusion 3 Medium release, which the company acknowledges didn’t meet expectations.

“This release didn’t fully meet our standards or our communities’ expectations,” Stability AI stated.

Rather than rushing a quick fix, Stability AI says it invested time in developing a more robust solution.

The flagship model, Stable Diffusion 3.5 Large, boasts 8 billion parameters and operates at 1 megapixel resolution—making it the most powerful in the Stable Diffusion family. Alongside it, the Large Turbo variant offers comparable quality but generates images in just four steps, significantly reducing processing time.

A Medium version, scheduled for release on 29th October, will feature 2.5 billion parameters and support image generation between 0.25 and 2 megapixel resolution. This variant is specifically optimised for consumer hardware.

Benchmark comparing the performance of the new Stable Diffusion 3.5 image generation models from Stability AI.

The models incorporate Query-Key Normalisation in transformer blocks, enhancing training stability and simplifying fine-tuning processes. However, this flexibility comes with trade-offs, including greater variation in outputs from identical prompts with different seeds.

Stability AI has implemented a notably permissive community licence for the release. The models are free for non-commercial use and available to businesses with annual revenues under $1 million. Enterprises exceeding this threshold must secure separate licensing arrangements.

The company emphasised its commitment to responsible AI development, implementing safety measures from the early stages. Additional features, including ControlNets for advanced control features, are planned for release following the Medium model’s launch.

Stability AI’s latest image generation models are currently available via Hugging Face and GitHub, with additional access through platforms including the Stability AI API, Replicate, ComfyUI, and DeepInfra.

(Image Credit: Stability AI)

See also: Anthropic unveils new Claude AI models and ‘computer control’

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

Tags: ai, artificial intelligence, image generation, models, open source, open-source, stability ai, stable diffusion

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Share post:

spot_imgspot_img

Popular

More like this
Related

Alibaba Marco-o1: Advancing LLM reasoning capabilities

Alibaba has announced Marco-o1, a large language model (LLM) designed to tackle both conventional and open-ended problem-solving tasks.

New AI training techniques aim to overcome current challenges

OpenAI and other leading AI companies are developing new training techniques to overcome limitations of current methods. Addressing unexpected delays and complications in the development of larger, more powerful language models, these fresh techniques focus on human-like behaviour to teach algorithms to ‘think.

Ai2 OLMo 2: Raising the bar for open language models

Ai2 is releasing OLMo 2, a family of open-source language models that advances the democratisation of AI and narrows the gap between open and proprietary solutions.

Generative AI use soars among brits, but is it sustainable?

A survey by CloudNine PR shows that 83% of UK adults are aware of generative AI tools, and 45% of those familiar with them want companies to be transparent about the environmental costs associated with the technologies.