2 min Applications

Alibaba beats OpenAI and Anthropic in AI tests

Alibaba beats OpenAI and Anthropic in AI tests

Alibaba Group Holding published benchmark scores and praised what it called itself the world’s leading performance with the release of its new model for AI.

This was reported by news agency Bloomberg. The upgraded Qwen 2.5 Max edition outperformed Meta Platforms’ Llama model and DeepSeek’s V3 model in several tests, according to figures from Alibaba Cloud’s announcement on WeChat. Along with Tencent Holdings and Baidu, Alibaba has invested significant resources in its cloud services segment and is engaged in a fierce battle to attract Chinese AI developers to use its tools.

Alibaba uses DeepSeek as benchmark

DeepSeek, a 20-month-old startup from Alibaba’s hometown of Hangzhou, became a global sensation this week and is playing a prominent role as the first benchmark against which Alibaba now seems to measure itself. Alibaba Cloud also shared scores showing that its AI outperforms certain benchmarks compared to models from OpenAI and Anthropic.

Price war between cloud service providers

Cloud service providers such as Alibaba and Tencent have lowered their prices in recent months in an effort to attract more users. DeepSeek has already contributed to that price war, along with half a dozen other promising AI startups in China that have obtained funding at unicorn valuations.

Also read: Tech stocks sink due to DeepSeek, but a rebound is in sight