Koyeb
Serverless cloud platform for AI inference and applications
Koyeb is a serverless infrastructure provider founded in 2021 by former Scaleway executives. It offers global deployment for AI workloads on CPUs, GPUs, and Tenstorrent accelerators, aiming to simplify scaling for developers.
- Founded in 2021 by former Scaleway executives
- Offers serverless access to Tenstorrent AI accelerators
- Supports automatic scaling from zero to hundreds of servers
- Backed by Samsung Next, MongoDB, and Isai
- Provides global deployment for CPU and GPU workloads
What it is
Koyeb operates as a high-performance serverless platform designed for intensive computing tasks. The company enables teams to run generative AI inference, manage agent fleets, and host APIs without handling underlying hardware. Founded in 2021, it provides a unified interface for deploying applications across central processing units, graphics processors, and specialized accelerators worldwide. Its infrastructure supports automatic scaling from zero to hundreds of servers, ensuring cost efficiency and low latency for production-grade workloads.
How it came about
The venture began in 2021, established by Yann Leger alongside colleagues Edouard and Bastien. These founders previously spent a decade building large-scale cloud systems at Scaleway. They observed that while basic CPU hosting had become a commodity, advanced platforms remained costly and complex to operate. Their goal was to create a next-generation cloud experience that combined seamless deployment with powerful hardware. The team has since focused on adapting their infrastructure to meet the specific demands of generative AI and agentic workflows.
The compute
The platform abstracts infrastructure management, allowing users to deploy via command line or git push. It supports Docker containers and multiple programming languages. In early 2025, Koyeb integrated Tenstorrent AI accelerators, offering instances like the TT-N300S and TT-LoudBox. These provide access to RISC-V based processors and open source tools such as TT-Metalium. The system also runs Nvidia GPUs and standard CPUs, featuring sub-second scale-to-zero capabilities and sandboxes for secure code execution.
Where the GPUs are
Koyeb maintains a globally distributed network of data centres. This geographic spread allows applications to be deployed in seconds across various regions. The company partners with hardware providers to host its diverse compute resources, including PCIe boards from Tenstorrent. This global footprint ensures low-latency connections for users regardless of their location, supporting the rapid scaling required by modern AI applications.
Customers and deals
The service targets developers and businesses needing to scale AI-intensive applications. Users range from those running simple inference tasks to teams managing large fleets of autonomous agents. The platform is designed for production environments where reliability and speed are critical. By removing the burden of infrastructure maintenance, it appeals to engineering teams who wish to focus on application logic rather than server management.
Money and ownership
Investment comes from a group of prominent backers including Serena, Isai, and Samsung Next. Other supporters include MongoDB, Acequia Capital, Kima Ventures, and Plug & Play Ventures. The company remains privately held and continues to raise capital to expand its hardware offerings and global reach. Financial details regarding valuation or revenue have not been disclosed in the available records.
Its place in AI
Koyeb positions itself as a hardware-agnostic alternative in the cloud market. By supporting non-Nvidia chips like Tenstorrent, it contributes to the diversification of AI infrastructure. The company draws on its founders' experience with ARM server adoption at Scaleway to promote open ecosystems. It participates in industry events such as NVIDIA GTC and AI Engineer Paris, highlighting its role in the evolving landscape of serverless AI deployment.
Questions people ask about Koyeb
What type of hardware does Koyeb support?
Koyeb offers a mix of computing resources, including standard CPUs, Nvidia GPUs, and Tenstorrent AI accelerators. The platform recently added support for Tenstorrent's RISC-V based chips, providing developers with alternatives to traditional graphics processing units.
How does the scaling mechanism work?
The platform uses serverless technology to automatically adjust resources based on demand. It can scale an application up to hundreds of servers during traffic spikes and scale down to zero when idle, helping users control costs while maintaining performance.
Who are the main investors behind Koyeb?
The company is backed by several notable firms, including Samsung Next, MongoDB, Isai, Acequia Capital, Kima Ventures, and Plug & Play Ventures. These investors support its mission to simplify AI infrastructure for developers.
Good reading and sources
There is no English Wikipedia article on Koyeb yet, so the facts on this page come from the sources listed above: koyeb.com/about, techcrunch.com/2025/02/25/serverless-cloud-platform-koyeb-lets-develop. Some are the subject's own pages, so its claims about itself are reported as its claims. EveryNeocloud wrote the text in its own words, with AI assistance, and compares it with the sources to make sure no passage is copied.
It may contain mistakes, and figures change over time. Nothing here is investment or buying advice. If something is wrong, tell us. Last refreshed 2026-11-14.