#HKSTARTUPS 阿里巴巴最新開源語言模型 Qwen3-235B-A22B-2507 在多項基準測試中擊敗 Kimi-2 和 Claude Opus，表現令人矚目。此模型不僅提升推理與程式碼生成能力，更推出低資源版本 FP8，讓企業能以較低成本運行。開源、效能佳、可本地部署，Qwen 正迅速成為企業級 AI 的有力競爭者。
Alibaba’s newly released open-source LLM, Qwen3-235B-A22B-2507, outperforms Moonshot’s Kimi-2 and Anthropic’s Claude Opus in key benchmarks. With an FP8 variant enabling cheaper and faster inference, and a shift toward specialized instruct/reasoning models, Alibaba is positioning Qwen as a high-performance, enterprise-ready alternative to closed systems.
📌 一杯咖啡價錢連接 Web3 世界 https://patreon.com/wanszezit
Full article https://venturebeat.com/ai/alibabas-new-open-source-qwen3-235b-a22b-2507-beats-kimi-2-and-offers-low-compute-version/