Alibaba’s Qwen3 Coder gains ground on rivals
Alibaba’s new open source coding model, Qwen3 Coder, is beginning to reshape the competitive AI landscape, showing early signs of denting the market share of established players such as Anthropic and Google.
According to OpenRouter, which tracks requests routed to different AI systems, the model has moved up sharply since its late July release. In mid-July, Anthropic’s Claude Sonnet 4 commanded 46.3% of usage share on the platform, while Google’s Gemini 2.5 Pro accounted for 13.4%. By the week of August 11, Claude’s share had fallen to 32.3% and Gemini slipped to 5.8%. Over the same period, Qwen3 Coder climbed to 20.7%, reflecting fast adoption among developers.
“The most striking change, however, is the meteoric rise of Qwen: Qwen3 Coder, which has claimed a remarkable 20.7% of the market share in just a few weeks,” OpenRouter’s data observed.
Qwen3-Coder-480B-A35B-Instruct, the technical name for the model, is designed as a Mixture-of-Experts (MoE) system. It has 480 billion total parameters, with 35 billion active per forward pass (8 out of 160 experts). The model is built for tasks such as function calling, tool use, and long-context reasoning across code repositories.
Unlike its American counterparts, Qwen3 Coder is open source. This gives developers the ability to adapt it for specific applications, an advantage not possible with closed-source offerings. As noted in the analysis, “Developers thus likely believe they have more control over the model, as opposed to closed-source models, which are run entirely by large corporations, and as OpenAI recently showed, can be taken down at will.”
For Alibaba, the success of Qwen3 Coder comes at a time when its core e-commerce business is seeing slower growth. The model’s uptake signals progress in the company’s broader push into AI, particularly in open-source innovation where China is aiming to match or overtake Western rivals. “Even more interestingly, Qwen is made by a Chinese company, so a Chinese company is taking away market share from American labs,” the commentary highlighted.
Qwen3 Coder’s rise is part of a wider trend in China’s AI ecosystem, alongside models such as DeepSeek, Kimi K2, and Z.ai. Analysts suggest that the momentum around Alibaba’s model could strengthen the company’s positioning in AI and potentially influence its longer-term business prospects.
It remains to be seen whether Qwen3 Coder can sustain its early growth, but the initial figures indicate a notable shift in developer preference, marking one of the strongest open-source challenges yet to Google and Anthropic’s dominance in code generation.
