SGLang team with Google and Meta open-sources SGLang-TorchTPU
10/20/2026 — 10/21, 02:19·1 sources (all)·1 reports (all)
Key takeaways
- RadixArk's SGLang team, working with Google and Meta, has open-sourced SGLang-TorchTPU for production-grade model serving on TPUs in native…
Background & analysis
On October 20, 2026, the SGLang team at RadixArk announced the open-source release of SGLang-TorchTPU, built in collaboration with Google and Meta. According to the team's announcement, the stack is designed to run production-grade model serving on TPUs in a PyTorch-native way.
The launch lineup of supported models includes Muse Spark Open, GLM-5.3, GLM-5.3 Flash, Kimi K3 and DeepSeek, among others. The available material covers only the open-source release and this initial model list; no further progress has been disclosed.
AI-generated from 1 reports · updated 2 hours ago
Latest turnThe SGLang team at RadixArk, working with Google and Meta, has open-sourced SGLang-TorchTPU, a PyTorch-native serving stack for production workloads on TPUs. Launch models include GLM-5.3, GLM-5.3 Flash, Kimi K3 and DeepSeek models.
- Heat index
- 56
- All sources
- 1
- All reports
- 1
- First seen
- 20 hours ago
Reports on this story headlines open the original
The SGLang team at RadixArk, working with Google and Meta, has open-sourced SGLang-TorchTPU, a PyTorch-native serving stack for production workloads on TPUs. Launch models include GLM-5.3, GLM-5.3 Flash, Kimi K3 and DeepSeek models.
LMSYS / SGLangFirst-partyAI score 55
Other stories people are talking about
- Heat index 513NewSurgeGoogle and Meta open-source TorchTPU backend for TPUs6 sources
- Heat index 1022DHS orders federal civilian agencies to purge DeepSeek tech16 sources
- Heat index 978SurgeDeepSeek open-sources DeepInfer large-scale inference framework12 sources
- Heat index 791RisingMiniMax M3.1 Flash launches on Cline and gmi_cloud17 sources
- Heat index 526RisingNscale in talks with Anthropic over roughly $45B compute deal8 sources
How is heat calculated?About the methodHide
Heat counts how many independent sources covered a story in the last 48 hours: one source counts once no matter how many posts it published, decaying with a 24-hour half-life. What ranks first is what many people are talking about.
This page aggregates public feeds. Headlines and summaries are machine-organized and remain the property of the original authors; verify important facts at the source. If you believe a headline or summary infringes your rights, email the contact address in the footer with the page URL and basis of your claim, and we will remove or replace it after review.
- Surge
- Discussion rising fast
- New
- First report within 6 hours
- Rising
- Still gathering discussion