Alibaba’s new model promises Opus 4.6-level performance on your laptop
Alibaba recently made the open weights of its 2.4 trillion parameter Qwen3.8 model available. That’s a massive model, and its The post Alibaba’s new model promises Opus 4.6-level performance on your laptop appeared first on The New Stack .
Alibaba has recently released a dense 27 billion parameter version of the Qwen 3.8 model, which can be run locally on a well-equipped MacBook Pro or Mac Studio. According to Alibaba's benchmarks, this model's performance is in the same league as Anthropic's Opus 4.6 running at its Max setting, and it even outperforms Anthropic's former flagship model in certain tasks such as coding and knowledge work.
The model also has vision capabilities, including video processing, making it a compelling option for local usage. Alibaba's benchmarks describe the original checkpoint, not the quantized versions most local users will run, which may result in a quality tradeoff. However, even Google's mid-tier Gemini 3.6 Flash only scored 49% on the DeepSWE agentic coding benchmarks, which Qwen 3.8 27B surpassed with a score of 42.2.
To run the model locally, users will need a Mac with at least 32GB of unified memory for the 4-bit version or 48GB or 64GB for higher precision and longer prompts. The model supports a context window of 262,000 tokens, extendable to 1 million tokens using the YaRN method.
Written by urgent.news from The New Stack's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
