RadixArk Stock
| Valuation | $400M1 |
| Valuation Date | May 5, 2026 |
| Implied Valuation | n/a |
| Last Price | n/a |
| 52-Week High | n/a |
| 52-Week Low | n/a |
| YTD Return | n/a |
| 1-Year Return | n/a |
| 3-Year Return | n/a |
| Total Funding | $100M1 |
| Latest Round Amount | $100M1 |
| Latest Round | Seed |
| Funding Rounds | 1 |
| IPO Status | Unknown |
| Expected IPO Date | n/a |
| Revenue | n/a |
| Revenue Type | n/a |
| P/S Ratio | n/a |
About RadixArk
RadixArk is an AI infrastructure company building open source systems for training, deploying and running large AI models. It is the company behind SGLang, an open source inference engine that serves trillions of tokens a day for companies like Google, Microsoft, NVIDIA and xAI. It also maintains Miles, an open source framework for large-scale reinforcement learning and post-training. RadixArk makes money by selling managed infrastructure and tooling built on top of these projects, which developers, enterprises and research labs use to train, fine-tune and serve models at scale. The company was founded in 2025 by SGLang creators Ying Sheng and Banghua Zhu, who previously built AI systems at xAI and NVIDIA.
Invest in RadixArk
RadixArk is a private company, so you can't purchase it with a regular brokerage account. However, the company is listed on Forge.
Forge is a trusted platform for transactions of shares of private companies. It is a subsidiary of The Charles Schwab Corporation.
Invest on ForgeDisclosure: Stock Analysis may be compensated for user sign-ups through Forge links.
Related Companies
News
RadixArk Releases Miles v0.1 Open-Source RL Framework
RadixArk released Miles v0.1, a production-oriented reinforcement-learning stack for frontier LLM post-training that combines SGLang rollouts with Megatron-LM or FSDP2 training. The framework targets ...
RadixArk joins Google to bring full SGLang features to TPUs
RadixArk and Google Cloud partnered to bring SGLang inference to Google TPUs, enabling high-throughput serving with Radix Cache, speculative decoding, and a planned PyTorch-native SGL-torchtpu backend...
SGLang and Miles Add Day-0 Support for Inkling, a Frontier Multimodal Model
RadixArk's SGLang and Miles add day-0 inference and reinforcement learning support for Thinking Machines Lab's Inkling multimodal model, with dedicated optimizations for its architecture.
RadixArk serves DeepSeek V4 at 3.5K+ tokens/sec/GPU on DigitalOcean and AMD
DigitalOcean details how RadixArk brought DeepSeek V4 to production on AMD Instinct MI350X GPUs with SGLang, achieving roughly 10x throughput gains through joint engineering with AMD.
Can Miles Make Large-Scale LLM RL Post-Training Practical for the Enterprise?
Futurum Group analyzes RadixArk's Miles framework and its potential to lower the infrastructure burden of large-scale reinforcement learning post-training for enterprises.
Miles: A PyTorch-Native Stack for Large-Scale LLM RL Post-Training
The PyTorch Foundation blog features RadixArk's Miles framework, which composes SGLang, NVIDIA Megatron-LM and Ray into a PyTorch-native stack for frontier-scale reinforcement learning.
RadixArk Launches with $100 Million in Seed Funding Led by Accel
RadixArk launched with $100 million in seed funding at a $400 million post-money valuation, led by Accel and Spark Capital with participation from NVentures, AMD, and other investors.
RadixArk Launches with $100 Million in Seed Funding Led by Accel to Grow SGLang and Democratize Frontier AI Infrastructure
RadixArk launched with $100 million in seed funding at a $400 million post-money valuation, led by Accel and co-led by Spark Capital, to commercialize the SGLang open-source AI inference engine.
Our $100M Seed to Build Open Infrastructure for Frontier AI
RadixArk officially launches with $100 million in seed funding at a $400 million post-money valuation, led by Accel and co-led by Spark Capital, to grow SGLang and Miles and build managed AI infrastru...
Nvidia, AMD back $100M round for AI tooling startup RadixArk
SiliconANGLE reports on RadixArk's $100 million seed round, noting backing from Nvidia's NVentures, AMD, Databricks and Broadcom CEO Hock Tan, and plans for managed infrastructure built on SGLang and ...
Our Investment in RadixArk: Building the Open Infrastructure for AI
Accel explains its lead investment in RadixArk, highlighting SGLang's deployment across hundreds of thousands of GPUs and the team's track record with open-source projects like Vicuna and FastChat.
Introducing Miles: A Production-Ready RL Framework for Large-Scale MoE Training
RadixArk open-sources Miles, a reinforcement learning framework for post-training large language models that pairs SGLang rollout with Megatron-LM training.