Zyphra announced the launch of Zyphra Cloud, a full-stack artificial intelligence platform powered by AMD (NASDAQ: AMD) Instinct MI355X GPUs on TensorWave's infrastructure. The platform combines model serving, agent infrastructure, and scalable compute into a single system for building and deploying AI applications.

The platform launches with Zyphra Inference, a serverless inference service that provides access to open-weight models including DeepSeek V3.2, Kimi K2.6, and GLM 5.1. The service targets production-grade applications such as agentic coding, deep research, and workflow automation through custom kernels, long-context inference algorithms, and advanced parallelism schemes.

"Zyphra Cloud is the natural extension of our research. We've spent years building, optimizing, and validating AI systems on AMD infrastructure, and are now bringing that capability to market as a platform for developers and enterprises," said Krithik Puthalath, founder and CEO of Zyphra.

Negin Oliver, corporate vice president of business development for AI at AMD, stated that the collaboration demonstrates how optimized AI software combined with AMD's accelerator architecture can deliver AI inference performance for demanding open-weight models.

The San Francisco-based company plans to expand the platform beyond inference to include distributed post-training services such as reinforcement learning and fine-tuning, sandboxed agent environments powered by AMD EPYC CPUs, and access to dedicated GPU clusters.

TensorWave, backed by AMD Ventures and Magnetar, provides the infrastructure for the platform and is among the first cloud providers to deploy AMD Instinct MI355X GPUs. Zyphra Cloud is available immediately through the company's website.