Cast AI

Kimchi Coding is now GA! After Early Access, Kimchi is ready for enterprise production: an autonomous multi-model coding agent that delivers frontier-quality AI coding at 2.5x lower cost, with full data sovereignty built in from day one. The premise is simple: the future of AI coding isn't a single frontier model. It's an orchestration layer that routes each task to the right model at the right cost. Kimchi is built for exactly that. Most work goes to open-weight models. The hardest tasks go to the most capable ones. Budget governance, data sovereignty, and cost observability ship in the product, not as add-ons. The results speak for themselves. In shadow-mode evaluations against a commercial-models-only baseline, Kimchi delivers 2.5x lower cost while matching or exceeding quality on spec-match and test-pass rates. The era of paying frontier prices for every line of code is over. Read more in the first comment.

Most companies are scaling AI workloads without a playbook for the infrastructure costs that come with them. Laurent Gil Founder and President of Cast AI, and Murat Kilicoglu, Partner at Cota Capital, are sitting down to talk about exactly that: how GPU markets are evolving, why AI infrastructure costs are under the microscope, and what automation has to do with whether your AI investment pays off. Register now: https://lnkd.in/gkNF86A6

We have officially launched our India Region in Mumbai 🇮🇳 Indian enterprises and global organizations running AI workloads in India can now operate the full Cast AI platform on local infrastructure with complete data residency and in-country support from day one. This isn't a limited preview. The Mumbai region launches with full feature parity, on the same global release schedule, built on the same highly available architecture that powers Cast AI deployments worldwide. This launch builds on our Bengaluru office, which opened in late 2024. India is a long-term commitment, not a checkbox. Read more in the first comment.

If you're heading to KubeCon India this Wednesday - don't miss these two talks at our booth: 14:30 - Scaling EdTech at Hyperspeed Avinash Gupta (Sr. Director of Engineering, PW (PhysicsWallah) + Gaurang Singh (Regional Field CTO APAC, Cast AI) on how PhysicsWallah cut Kubernetes costs without slowing down. Room 205 AB. 16:30 - Cut the Bill, Not the Brains Matas Kaminskas (Sr. Engineering Manager, Cast AI) on AI coding agents that ship more and cost less. Room 205 AB. And before or after - find us at Booth P5 for a chat, a demo, and some swag 🇮🇳

Clusters & Cocktails Mumbai is this week - a few spots still left. June 18, 7 PM, Baglami BKC. Grab yours → https://lnkd.in/dhj4UGKh

Kimchi Coding, powered by Cast AI, is the first autonomous coding agent to offer MiniMax M3 - available now through our Early Access program. M3 scored 59% on SWE-bench Pro, ahead of several leading commercial models. Inside Kimchi Coding's multi-model orchestration layer, it becomes the default builder model, routed intelligently alongside lighter open-weight models based on task complexity and cost. Kimchi is obsessed with token optimization. Its orchestration engine routes each task to the best-fit model, scores generated code, and runs continuous feedback loops, so output stays accurate while token use is minimized at every step. That's what makes it both more accurate and more cost-efficient than relying on a single commercial model. The result: 2.5× lower cost than a commercial-models-only baseline, with matching or better output quality. The link to the full press release is in the first comment 👇

FinOps X 2026 was a milestone week for the industry and for Cast AI! Our cofounder and president Laurent Gil was in the room when the Tokenomics Foundation became official. He sat down with theCUBE to talk about what tokenomics actually means for infrastructure teams GPU economics, AI spend governance, and where this is all heading. Worth a watch. Our own Esmira Bayramova also took the stage, covering GPU utilization data, token economics, and what teams running AI workloads at scale need to start thinking about now. The conversation at FinOps X made one thing clear: the "figure out cost later" era is over. The FinOps community is ready to own this. We had a booth on the floor too and the energy from practitioners was exactly what you'd hope for at a moment like this. Thank you San Diego. See you at Tokenomicon! 👋

Don't miss out - register today! 👇

KubeCon India is almost here - and we're throwing an after party 🎉 Clusters & Cocktails Mumbai is happening June 18, 7 PM at Baglami, BKC. Big thanks to our partners making it happen - Chainguard, Datafy.io, Excelien, and Amazon Web Services (AWS). Spots are limited, grab yours → https://lnkd.in/e7fyfK3p