The most interesting point in the comments
Alibaba releases new artificial intelligence model Qwen3.8-Flash – An affordable option designed to incentivize deployment of the Chinese developer’s flagship model.
Image source: qwen.ai
The company has released open weights for the Qwen3.8-Flash model on the Hugging Face and ModelScope platforms. Its scale is 125 billion parameters; Alibaba said that in terms of work quality, it has reached the level of the world’s leading systems, including Anthropic Opus 4.6 and DeepSeek V4-Flash. Compared with Qwen3.7-Plus, the training and inference costs of the new Qwen3.8-Flash are nine times cheaper. It has a new generation architecture and will be used in the Qwen 4 series models.
The size of the upper and lower windows of Alibaba Qwen3.8-Flash is 1 million tokens. Developers will open access to it via API – 1 million input tokens cost $0.16 and 1 million output tokens cost $0.47.
The e-commerce giant has become a major player in the AI industry. The company speeds up the research and development of new models. In early August, Alibaba released its updated flagship Qwen3.8-Max. The company also today launched an international version of its enterprise artificial intelligence agency QwenWork to expand sales in global markets such as Asia, Latin America and the Middle East.
Tell Google to receive more frequent links to our news about artificial intelligence
If you find an error, select it with your mouse and press CTRL+ENTER.
