Alibaba Drops Biggest AI Model as DeepSeek Cuts Costs

The release of Qwen3.8-Max marks Alibaba's most ambitious AI push yet, with the model reportedly outperforming its predecessors on a range of benchmarks, including complex reasoning and multilingual tasks. The company said the model is designed for enterprise use, targeting sectors such as finance, healthcare, and manufacturing.

DeepSeek, a rising startup, countered with V4-Flash, which it claims delivers comparable performance at a fraction of the cost. The company said the model's training and inference costs are significantly lower than existing alternatives, a move that could pressure larger players to rethink their pricing strategies.

The two launches come as Chinese tech firms race to close the gap with US leaders like OpenAI and Google. Beijing has encouraged domestic AI development, and the latest models show Chinese companies are now competing not just on capability but also on cost efficiency.

Industry analysts note that DeepSeek's aggressive pricing could democratize access to advanced AI in China, where many small and medium businesses have been priced out of the market. Alibaba, meanwhile, is betting that its scale and cloud infrastructure will give it an edge in serving large enterprises.

Both companies are expected to release technical papers and open-source components in the coming weeks, a move that could further accelerate innovation and competition in the sector.