Alibaba's new model beats its predecessor at one-ninth the training cost
Qwen3.8-Flash-Next has 125 billion parameters but activates only 6 billion per token. It outperforms a model three times its active size.
Models
Alibaba
Qwen's new model activates just 6 billion of its 125 billion parameters
Alibaba says it beats the larger previous model at roughly one-ninth the training cost. The release is framed as an architecture preview of Qwen4.
Models
Apple trained its China-specific model with help from Alibaba
Apple is reported to have developed a large language model specific to the Chinese market together with Alibaba, and has registered the service with China's regulator.
Companies
Alibaba releases Qwen 3.8 models with open weights
Alibaba's Qwen team has published the 27-billion-parameter multimodal Qwen3.8-27B under the Apache 2.0 licence. The model handles a 262,000-token context natively.
Models
Alibaba's Qwen Models Surpass 3 Billion Downloads in 6 Months
Alibaba's open-weight Qwen model family surpassed 3 billion downloads globally in six months, outpacing Google and Meta. According to Hugging Face data, Qwen has become the most-derived model in the open-source ecosystem.
Models