Alibaba unveils QwQ-32B (Qwen2.5-32B), a new AI reasoning model that competes with DeepSeek’s R1

Just two weeks after unveiling the Qwen 2.5-Max model, Alibaba is back with another AI reasoning model, QwQ-32B (Qwen2.5-32B), which it claims can go head-to-head with DeepSeek’s R1.

Announcing the launch on X, Alibaba described QwQ-32B as a reasoning model designed to match the performance of leading alternatives like DeepSeek-R1 while operating with significantly fewer parameters.

Investors seem to like what they’re hearing—Alibaba’s Hong Kong-listed shares jumped 8.39% on Thursday, hitting a 52-week high, while its U.S.-traded stock saw a 2.5% bump in premarket trading. Since the start of the year, Alibaba shares in Hong Kong have surged nearly 71%.

QwQ-32B vs. DeepSeek R1: Can Alibaba’s AI Model Compete?

QwQ-32B runs on 32 billion parameters, a fraction of DeepSeek’s 671 billion. But Alibaba says it achieves comparable results, thanks to a more efficient design. Only 37 billion parameters are active during inference, making it a leaner model while still delivering a strong performance. AI developers are increasingly prioritizing efficiency, aiming for models that require less computing power without sacrificing capability.

Today, we release QwQ-32B, our new reasoning model with only 32 billion parameters that rivals cutting-edge reasoning model, e.g., DeepSeek-R1.

Blog: https://t.co/zCgACNdodj
HF: https://t.co/pfjZygOiyQ
ModelScope: https://t.co/hcfOD8wSLa
Demo: https://t.co/DxWPzAg6g8
Qwen Chat:… pic.twitter.com/kfvbNgNucW

— Qwen (@Alibaba_Qwen) March 5, 2025

The company also highlighted its ongoing efforts to improve reinforcement learning (RL) techniques, stating that these advancements have led to notable improvements in areas like mathematics and coding. Alibaba believes that by continuously scaling RL, medium-sized models can reach performance levels previously seen only in much larger architectures. Along with the announcement, Alibaba shared links to the model’s documentation, demo, and chat interface, encouraging users to test it and provide feedback.

Alibaba says QwQ-32B is showing strong results, particularly in math and coding. The company expects to refine the model further, keeping pace with the growing competition in AI. The timing is no coincidence—DeepSeek’s R1, introduced earlier this year, set a new benchmark in reasoning models. Since then, established tech giants and AI startups alike have been scrambling to push their own advancements.

Chinese firms have been ramping up AI investments, with Alibaba positioning itself as a key player. The company has been pouring resources into its Cloud Intelligence unit, which played a major role in its recent profit growth. CEO Eddie Wu has made it clear that AI will be a core focus, saying, “Looking ahead, revenue growth at Cloud Intelligence Group driven by AI will continue to accelerate.”

Analysts are optimistic that AI could fuel Alibaba’s stock momentum. Bernstein analysts see a path for sustained earnings growth, while industry experts highlight how AI models are becoming more accessible. Futurum Group CEO Dan Newman noted on CNBC that DeepSeek’s emergence forced the industry to reconsider whether OpenAI would remain the dominant player or if challengers like Alibaba, Microsoft, and Google would take the lead.

Newman pointed out that AI models are moving toward commoditization, with companies working to bring down costs and make them more widely available. “As we see this more efficiency, this cost coming down, we’re also going to see use going off,” he said. While Nvidia has benefited from the AI training boom, he believes inference—how AI is actually applied in real-world use cases—is where the next wave of growth will happen.

Alibaba is betting on that shift. With QwQ-32B, the company is pushing forward in a space where efficiency and scalability are becoming just as important as raw model size. Whether it can truly compete with DeepSeek’s R1 remains to be seen, but one thing is clear: the AI race is only getting more intense.

Alibaba’s AI Push Extends Beyond Its Own Models

Alibaba isn’t just focused on building its own AI—it’s backing other Chinese AI startups as well. In late 2023, the company joined Tencent and other investors to invest $340 million in Zhipu, a rising competitor to OpenAI.

Zhipu has positioned itself as one of China’s top generative AI startups, playing a key role in shaping the country’s AI landscape. Meanwhile, tech giants like Alibaba, Tencent, and Baidu aren’t just investing—they’re actively developing their own AI models and rolling out new products powered by this technology.