16.2 C
Tuesday, July 16, 2024

OctoML Unveils Solution for Faster Image Generation

Must read

Industry-first architecture offers unmatched customization capabilities for image-generation applications.

OctoML launched “OctoAI Image Gen,” the industry’s first architecture for customizing image generation applications built on popular models, including Stable Diffusion. The new solution enables developers to dynamically apply thousands of customization assets to their image generation model via a single API within OctoAI. Now, businesses can deliver highly customized image generation applications at scale without sacrificing performance, speed and cost.

“Image generation applications have quickly gone from fad to real business, with many ecommerce, entertainment and creative organizations looking to differentiate their service with AI,” said Luis Ceze, CEO, OctoML. “But building these custom experiences with Stable Diffusion today requires engineering that simply doesn’t scale. Our new solution allows you to easily customize Stable Diffusion and tap into innovations in the ecosystem over a single API endpoint.”

Foundational to the new release is OctoAI’s new “Asset Orchestrator”, which enables developers to dynamically apply fine-tuning assets—like LoRAs and Checkpoints—to deliver powerful customization capabilities. The Asset Orchestrator also enables developers to access preloaded assets, import them from communities such as HuggingFace, or create their own via native fine-tuning within OctoAI. The solution then automatically applies the assets to the application as needed. The result is flexible fine tuning, without the monetary and performance costs of starting from scratch and maintaining a custom endpoint.

Benefits and features:

  • Unparalleled Speed and Cost Efficiency: Achieve accelerated image generation at just 2.8 seconds on SDXL. Leverage one API to perform infinite fine tuning customizations. Maximize value with low operational costs.
  • Advanced Model and Asset Management: Gain seamless access to the latest models upon release. Import and manage fine tuning assets from platforms like Civitai or HuggingFace. Personalize with native-fine tuning to maintain product integrity or replicate brand style.
  • Robust and Consistent Delivery: Asset Orchestrator ensures consistent, high-quality image results. Reliably handle 10x usage surges with proven performance.

Already a number of emerging businesses, including Storytime AI, Nightcafe and CALA, are using OctoAI’s image generation solution to power their applications.

“Our top priority is to deliver kid-safe, consistent, engaging images for our custom children’s stories,” said Brian Carlson, CEO, Storytime AI. “Previously, this process relied on heavy-handed prompt engineering. But OctoAI helped us stand up a whole new image gen architecture utilizing assets like LoRAs and checkpoints to create consistent visuals without the added complexity of prompt engineering.”

“Speed is key to the AI art experience we deliver,” said Angus Russel, Founder, NightCafe. “With OctoAI’s Image Gen Solution, we’ve increased our image generation speeds by 5X. The UX improvement from these low latency inferences has resulted in even more usage and growth for our platform.”

“OctoAI’s integration has been instrumental in making it possible for CALA to power the ability for our customers to fine-tune their image generation,” said Dylan Pyle, Chief Technology Officer, CALA. OctoAI has allowed us to accelerate our development and time to market with these new features while eliminating the typical costs that we would have faced by running multiple parallel model variants.”

More articles

Latest posts