Runware logo

Meet the startup simplifying image and video generation with Google Cloud

Google Cloud results

Runware built a state-of-the-art API to rapidly deliver Google Cloud's market-leading image and video generation models at scale.

Meeting user demand with "Day 0" access
to the latest AI models

Google Cloud models really are state of the art, and there's a significant step up in quality with every major release. I can't stress enough the importance of the information we get from the Google for Startups team in ensuring we release the best models as soon as they're available.

Ioana Hreninciuc

Co-founder, Runware

For developers and creative companies, the challenge of building powerful image and video generation features into their apps can be overwhelming. They need to source the best models from multiple providers, manage complex infrastructure for each one, and constantly keep up to date with the newest releases. Additionally, customers need support with context engineering, deeper platform integration, and front-end workflow deployment. This time-consuming process is typically too expensive to scale, and requires deep, specialized expertise of the latest models.

Ioana Hreninciuc and Flaviu Rădulescu founded Runware to eliminate this complexity. With its single, powerful API to handle all the logistics and infrastructure—including the different model standards, the hardware, and the scaling—Runware allows companies to offer comprehensive video and image generation capabilities with none of the complexity, to provide the best possible experience for their users.

Key to Runware's success is ensuring its customers always have access to the newest, most effective models as soon as they're released. The world of generative AI moves quickly, and the release of powerful new models can instantly change what's possible. Waiting weeks for access means developers risk being left behind. Runware needed a provider that wasn't only building the best models but would also work closely with the company to ensure developers had access immediately upon release.

As a participant in the Google for Startups program, Runware benefited from resources and credits, helping the company offset the high costs of testing and deploying new models as it scaled to meet customer demand. A key part of this relationship was the collaboration with the Google Cloud team to gain the latest information and technical guidance needed to integrate new models to be ready when they launch. This close alignment means Runware can make the newest, most advanced models from Google Cloud available as soon as they hit the market, helping its customers to offer users cutting-edge capabilities with the latest Google AI models.

"Google Cloud models really are state of the art, and there's a significant step up in quality with every major release," explains Ioana Hreninciuc, co-founder of Runware. "I can't stress enough the importance of the information we get from the Google for Startups team in ensuring we release the best models as soon as they're available."

Runware homepage featuring one API for all AI text

Redefining what's possible with
pioneering gen AI capabilities

By integrating models like Gemini, Imagen, including the Nano Banana family, and Veo 3, Runware centralizes access to Google's most effective models through a single API. This unified solution means Runware's customers can rapidly swap and combine models within their product toolsets, ensuring end users can always access the latest and most powerful models in a simple, cost-effective way.

Runware's decision to offer Google Cloud models is driven by the need to ensure its customers can provide the highest quality image and video generation capabilities to end users. Imagen's Nano Banana model, for instance, keeps details consistent across different generations, such as maintaining the look of a person or combining multiple items into one photorealistic image.

Nano Banana was the first model to deliver this degree of quality, which is why it's been so widely adopted. In terms of the sheer quality of the result, Google AI models are currently the best in the industry.

Ioana Hreninciuc

Co-founder, Runware

Grid of AI model thumbnails on Runware library page

This is key for end users engaged in high-volume commercial content creation, such as ecommerce product photos or iterative advertising campaigns.

"Nano Banana allows users to make really precise changes and delivers very high-quality image edits, which are important to users," explains Hreninciuc. "Nano Banana was the first model to deliver this degree of quality, which is why it's been so widely adopted. In terms of the sheer quality of the result, Google AI models are currently the best in the industry."

Similarly, Veo 3 introduced realistic audio capabilities alongside video generation, which is a major requirement for creators making social media content or ads. What's more, because Google Cloud models are adept at understanding user intent, developers can achieve the result they want in fewer attempts. A task that once required many costly iterations might now take only a few, meaning these models save users time, effort, and resources in the long run. Veo 3 and Nano Banana are the most popular models on Runware's platform due to this combination of power and ease of use.

"Google Cloud models are very good with reasoning, which means they understand what the user is trying to do," Hreninciuc explains. "Most people don't really know how to prompt the model to get what they need. Google Cloud models solve this problem by understanding what the user is trying to achieve, which can be the difference between 20 iterations and just two, making them far more cost effective."

Runware Playground UI showing AI green eye edit

Expanding the model portfolio
with Google Cloud

Our aim is to simplify the user experience by providing everything in one place. As we continue to expand our offering, Google Cloud models will remain a key component of our portfolio. They will be the flagship models for every modality that we offer.

Ioana Hreninciuc

Co-founder, Runware

Runware's ability to provide immediate access to popular Google models Nano Banana and Veo has been a core driver of the platform's success, with improvements in quality driving users to consume more content. This increased demand has helped to fuel Runware's growth by over 40x in a year, with over 2.5 billion assets generated to date.

As it grows, Runware is expanding its offering to provide a complete toolkit for all generative AI, including integrating Google Cloud large language models (LLMs), audio models, and even models for 3D generation. This expansion is designed to help developers work even faster with even more impressive results. For instance, a client building a website generator will soon be able to generate both the images, with Imagen, and the accompanying text, with Gemini, through a single, unified Runware platform.

"Our aim is to simplify the user experience by providing everything in one place," Hreninciuc concludes. "As we continue to expand our offering, Google Cloud models will remain a key component of our portfolio. They will be the flagship models for every modality that we offer."

Runware founders standing in front of its logo

Runware gives developers instant access to video and image models to supercharge their creative projects. With its powerful, unified API, it puts the whole model ecosystem at developers' fingertips.

Industries: Technology, Media

Location: UK

Products: Google Cloud, Gemini, Imagen, Nano Banana Flash, Nano Banana Pro, Veo