
Alibaba has unveiled its latest AI model, Qwen3.8-Max, claiming it rivals top US models like Anthropic's Fable 5. With 2.4 trillion parameters, the model is designed to compete at the highest levels, and Alibaba plans to release its weights next week, emphasizing openness. This move is part of a broader strategy by Chinese firms to close the gap with US AI leaders. The release highlights ongoing tensions in AI development and governance between China and the US.
Read originalEarlier coverage that leads up to this article, and what followed. Lines connect each piece to the closest one after it, converging here.
This release quietly expands llama.cpp's hardware support to include Qualcomm's Hexagon NPU on Linux arm64, a significant step for local inference on Snapdragon devices. It also updates CUDA builds to version 13.4 and introduces ROCm 10.0 binaries, keeping the project aligned with the latest NVIDIA and AMD driver ecosystems. KleidiAI on Apple Silicon is temporarily disabled in this build, likely due to stability checks rather than a feature rollback. For developers targeting edge AI or diverse GPU stacks, this update ensures broader compatibility without requiring custom compilation.
© Lev SelectorNVIDIA introduced the NVFP4 4-bit format and SoL-Pi technology, which uses 2x fewer tokens for improved efficiency.
Matt Wolfe · April 24, 2026 · Related
AI News · May 20, 2026 · Related
The Verge AI · July 20, 2026 · Same story
TechCrunch AI · July 20, 2026 · Related
AI News · July 21, 2026 · Related
WIRED AI · July 24, 2026 · Related
TechCrunch AI · July 26, 2026 · Related
The Verge AI · July 27, 2026 · Related
The AI Daily Brief · August 10, 2026 · Related
Hugging Face Blog · August 14, 2026 · Related
Sam Witteveen · August 18, 2026 · Related
Matt Wolfe · August 21, 2026 · Same story
The Rundown AI · August 27, 2026 · Background
Alibaba Announces Qwen3.8 Open-Weight Release
6 developments