Alibaba's Apsara Conference 2026

At Apsara Conference 2026, Alibaba announced a set of hardware, model, and product developments. The Zhenwu V900 chip, developed by T-Head, claims roughly 3x the performance of its predecessor the M890, with 216GB of on-package memory, 1.2 TB/s chip-to-chip interconnect via ICN Switch fabric, native FP8/FP4 support, and cluster support up to 500,000 cards. Mass production and commercial release are targeted for Q1 2027. All V900 performance claims are vendor-reported and have not been independently benchmarked. A successor chip, the J900, is already on the roadmap for Q3 2028. The predecessor M890, released May 2026, has shipped over 560,000 units to more than 400 external customers across 20 industries.

The Qwen team is pursuing Recursive Self-Improvement and planning a 5-10 trillion parameter model on the path to ASI. CEO Eddie Wu stated that machine-generated thinking is currently less than 3% of total human cognitive volume. Alibaba Cloud targets exceeding 20GW of global data-center capacity by 2032.

Qwen Intelligence launched with three mobile agents. The Mobile Planner Agent is claimed to rank first on MobilePA-Bench, MobilePA-Bench Business, and Memory benchmarks. The Mobile-Use Agent scored 82.1 on MobileWorld, 92.2 on MobileWorld-Real, 97.2 on AndroidDaily, and claims a 90% end-to-end success rate. The Mobile Creative Agent generates images in 3 seconds, described as approximately 2x faster than leading competitors. Public benchmarks, code, and leaderboards are available for testing the mobile-agent workflows. Qwen is open-sourcing a benchmark suite covering planning, cross-app execution, real-device performance, and safety. Reporting from November 2025 noted Alibaba deploying over 100 engineers to embed Qwen as a task-executing agent in Taobao and other retail apps, with a longer-term plan to monetize AI assistance in mobile shopping.

Qwen-Audio-3.1 upgrades ASR, TTS, and Realtime models and adds TTS-Next and ASR-Next. TTS-Next uses a unified language model plus diffusion framework to generate voice, sound effects, and background audio in a single pass. ASR-Next supports multi-speaker recognition with speaker labels, timestamps, emotion detection, and audio reasoning. Realtime mode supports simultaneous speaking and listening with interruption support and empathetic responses based on detected mood. Price cuts accompany the release: TTS by roughly 70%, Realtime by roughly 85%, and ASR by up to 95%.