Alibaba's Qoder Adds Linux ARM64 and LoongArch64 Builds, Plus UOS and Kylin Desktop Support
Alibaba's agentic coding platform Qoder now ships builds for Linux ARM64 and Loongson's LoongArch64 instruction set. It has also been adapted for UnionTech UOS Desktop V20 and Galaxy Kylin Desktop V10, with both the international and China editions updated on October 23.
Alibaba's agentic coding platform Qoder now ships builds for Linux ARM64 and for Loongson's LoongArch64 instruction set, and it has been adapted for UnionTech UOS Desktop V20 and Galaxy Kylin Desktop V10, according to Pandaily. Alibaba said the international and China editions both received the update on October 23.
What shipped
- Linux ARM64: a build covering Arm-based PCs and boards, released in the same train as the LoongArch build.
- LoongArch64: a build for Loongson's instruction set. Per Pandaily, Loongson's current chips use a fully self-developed instruction set that is not ARM64, not x86 and not licensed from either camp.
- Desktop adaptation: UnionTech UOS Desktop V20 and Galaxy Kylin Desktop V10.
The update brings the same demand-to-delivery agent workflow to desktops that many Chinese government and enterprise users already run, according to the report. On LoongArch machines, developers can call tools, edit files and run multi-agent jobs without switching to a foreign-architecture workstation.
Downloads are listed on Qoder's site, with Linux install notes that call out the new architectures.
China desktop footprint
With this release, Qoder's China desktop footprint covers HarmonyOS PC and HarmonyOS mobile clients, as well as UOS V20 and Kylin V10. For teams standardized on those systems, agent-assisted coding no longer requires a separate x86 or cloud-only environment.
Product background
Qoder launched globally in August 2025 as an agentic coding platform spanning a desktop IDE, CLI and JetBrains plugins. In May 2026, Alibaba Cloud moved it from an AI IDE toward an autonomous development workbench as Qoder 1.0, and folded the former Tongyi Lingma assistant into the China product line as Qoder CN.
The family now also includes a mobile client, QoderWake digital-employee tools and Qoder Cloud Agents.
Alibaba says that as of August 2026 the product had more than 6 million users and over 100,000 enterprise customers. Those adoption figures are the company's own and have not been independently verified.
No new model is part of this release. Model details, context window and pricing are not part of the announcement.
What this means
This is a distribution change, not a model change. The release extends where Qoder's agent workflow can run: first-party builds now exist for Arm-based Linux and for Loongson's own instruction set, alongside two named domestic Linux desktops.
For developers on UOS V20 or Kylin V10, the practical effect is that the agent tooling runs natively on the desktop they already use, rather than requiring an x86 or cloud-only environment. For LoongArch users specifically, the build means tool calls, file edits and multi-agent jobs can run on the machine itself.
The adoption numbers (6 million users, 100,000 enterprise customers) come from Alibaba and should be read as company claims. Whether the new builds match the feature set of the existing desktop clients has not been stated.
Related Articles
NVIDIA releases 96 GiB NVFP4 build of Qwen3.8-Flash-Next for single-GPU Blackwell and DGX Spark
NVIDIA has published Qwen3.8-Flash-Next-FP4-Local, a 96 GiB NVFP4 quantization of Alibaba's 125B-parameter MoE model, on Hugging Face. NVIDIA reports 91.5% on GPQA Diamond versus 92.2% for BF16. It requires NVIDIA's custom vLLM containers because upstream support has not landed.
OpenAI upgrades ChatGPT Chrome extension with multi-tab awareness and ChatGPT Work site control
OpenAI has upgraded its ChatGPT Chrome extension with a sidebar that reads across open tabs. Through ChatGPT Work, it can also control Chrome tabs to send messages, fill out forms and download files. TechRadar tested the form-filling feature and found it useful but in need of supervision.
Liquid AI launches RIPL, a real-time inference engine for LFM models, with 14.4 to 28.4x lower p95 latency
Liquid AI released RIPL, a real-time inference engine for its LFM models. On a single B200, LFM2.5-1.2B running on RIPL showed 14.4 to 28.4x lower p95 latency than Qwen3.5-0.8B running on vLLM.
NVIDIA expands Cosmos with world model training tools, simulation checkpoints and Data Factory blueprints
NVIDIA has expanded the Cosmos platform with new training tools, simulation checkpoints and Cosmos Data Factory blueprints. The company says Cosmos ranks in the top five on six different open model leaderboards.
Comments
Loading...