Cast AI launches Kimchi Coding for enterprise developers

Cast AI has launched the general availability of Kimchi Coding, an autonomous AI coding agent for software development teams.

The release moves Kimchi Coding out of early access and into production use. Akamai is already using the system.

Kimchi Coding is built around a multi-model approach rather than relying on a single large AI model for every task. The software routes work to different models based on complexity and cost, aiming to reduce spending while maintaining code quality.

Cast AI argues this reflects a broader shift among large organisations that want to reserve more advanced models for difficult tasks and use open-weight models for routine work. Kimchi Coding is designed as an orchestration layer for that process and can be deployed inside a customer's virtual private cloud or on dedicated Nvidia B300 graphics processing units managed through Cast AI's infrastructure.

In shadow-mode evaluations against a baseline using only commercial models, Kimchi Coding was 2.5 times cheaper while matching or exceeding quality on specification matching and test pass rates, according to Cast AI. The product also includes spending controls, from individual application programming interface keys to wider organisational budgets, along with automated shutdown of runaway agent loops and a dashboard that tracks costs by developer, team and project.

The product is closely tied to Cast AI's existing business in Kubernetes and graphics processing unit optimisation. Kimchi Coding uses the same underlying infrastructure engine as the company's wider platform, which it sells to businesses running cloud-native and AI workloads.

Cost focus

The launch comes as companies try to balance enthusiasm for AI coding tools with concern over the cost of using large proprietary models at scale. Governance, data location and spending oversight have become central issues for engineering leaders, particularly in regulated sectors and large multinational businesses.

Cast AI says Kimchi Coding addresses those concerns through hard spending caps, data sovereignty controls and self-hosted inference options. The product also has ISO 27001 and SOC 2 Type II certification and is compliant with General Data Protection Regulation requirements, according to the company.

Laurent Gil, president and co-founder of Cast AI, set out the company's position on the economics of the market.

"We built Kimchi to give every developer frontier-quality AI coding without frontier-sized bills or data risk," said Laurent Gil, president and co-founder of Cast AI. "General availability means that enterprises are now using Kimchi Coding at scale. The economics of AI coding are broken when you rely on a single commercial model. Kimchi fixes that, and does it without asking teams to compromise on quality, security, or control."

Early customer

Akamai is among the companies using the product, according to Cast AI. The networking and cloud services group is an example, the company says, of a large engineering organisation adopting a multi-model strategy for AI-assisted software development.

Dekel Shavit, senior director of engineering at Akamai, said the company saw the appeal of using different models for different jobs rather than treating one model as the answer to every task.

"Kimchi is fundamentally changing how our engineering team thinks about AI-assisted development," said Dekel Shavit, senior director of engineering at Akamai. "The multi-model approach keeps the quality of a single frontier model while cutting token costs dramatically. And the governance suite is the best we've seen in the market. Data sovereignty, cost observability, and budget control in one place. For an organization like Akamai, that makes a real difference."

Cast AI is known for software that helps companies manage Kubernetes environments and AI infrastructure more efficiently. It reached unicorn status after a strategic investment from Pacific Alliance Ventures, the corporate venture arm of Shinsegae Group, according to the company.

Its customer list includes BMW, Cisco, FICO, Hugging Face and Swisscom, reflecting a client base that spans industrial, telecoms and software sectors. With Kimchi Coding now generally available, Cast AI is extending that business into AI software development tools, where pricing pressure and governance demands are becoming as important as model performance.