Alibaba Unveils 2.4 Trillion Parameter AI Model Qwen3.8 with Multimodal Capabilities

• July 19, 2026

Alibaba has announced the latest iteration of its flagship artificial intelligence language model known as Qwen3.8. This new version significantly expands the model’s scale, now comprising an impressive 2.4 trillion parameters. The company highlighted the model’s multimodal processing capabilities, enabling it to handle diverse data types beyond text.

According to Alibaba’s internal assessments, Qwen3.8 demonstrates remarkable overall performance, ranking just below Anthropic’s Claude Fable 5, a notable competitor in the large-scale AI model domain. However, these performance claims have yet to be validated by third-party evaluations, as independent testing results have not been released to the public.

Advancing Model Scale and Versatility

The rapid growth in parameter count reflects Alibaba’s continued investment in pushing the boundaries of artificial intelligence technology. Models of this scale require significant computational resources and advanced training techniques, positioning Qwen3.8 among the largest language models developed to date.

Importantly, Qwen3.8’s multimodal capabilities mark a step forward in integrating different data types for comprehensive analysis and generation tasks, common in cutting-edge AI research. Such multimodality enables the model to process inputs that may include text, images, and potentially other forms of data, enhancing its versatility across varied applications.

In addition to unveiling the model, Alibaba announced plans to release Qwen3.8’s weights publicly. This move aligns with growing trends toward transparent AI development, potentially allowing researchers and developers worldwide to explore and build upon this technology. The details surrounding the timing and scope of this release were not disclosed.

Qwen3.8 forms part of Alibaba’s broader strategy to solidify its presence in the AI landscape through the creation of advanced generative models. As the competition among leading technology firms intensifies, innovations like Qwen3.8 could influence the development of future AI-driven products and services, particularly in natural language processing and multimodal applications.

While Alibaba’s announcement underscores the model’s potential, independent benchmarking will be crucial to validate its capabilities and compare them objectively with other major AI models in the industry.

Alibaba introduced Qwen3.8, a 2.4 trillion parameter AI model with multimodal features, positioned just behind Anthropic’s Claude Fable 5 on internal benchmarks.