Editorial illustration for Alibaba Tests Show Its New AI Model Rivals Top Competitors
Alibaba's Qwen3.8-Max Rivals Claude in New Tests
Alibaba Tests Show Its New AI Model Rivals Top Competitors
Alibaba put out a new model on Monday and said it's the biggest and most capable one the company has ever built. Qwen3.8-Max is now widely available to users, following a preview last month in which Alibaba described the system as "second only" to Anthropic's flagship, Claude Fable 5. The company put the model at 2.4 trillion parameters, a scale that lands it in the same conversation as the top systems out of Anthropic, OpenAI, and Moonshot AI, whose Kimi K3 has become a benchmark reference point for Chinese labs.
The timing matters. Washington and Silicon Valley have spent months arguing over how to handle the rapid pace of Chinese AI releases and whether the US can hold onto its technical lead. Another open, high-performing model out of Beijing adds fresh weight to that argument, particularly since Alibaba isn't just claiming parity in a press release. It backed the announcement with benchmark numbers and a public leaderboard ranking, the kind of evidence that's harder to wave off than a company blog post.
The open release of another highly capable Chinese AI model adds to sky-high tensions along multiple fronts in Silicon Valley and Washington over how to safely manage AI systems and retain the US’ technological edge over China.
Why this matters
Alibaba's own benchmarks are doing a lot of the talking here, and that's worth sitting with for a second. A company grading its own homework and then pointing to a crowdsourced leaderboard like Arena.AI isn't the same as independent, adversarial testing from developers who'll actually build on Qwen3.8-Max in production. For founders weighing which model to bet a product on, the gap between "matches Fable 5 on our tests" and "matches Fable 5 when you throw your weird edge cases at it" is exactly where real decisions get made.
What we do think matters: this is the third major model in weeks claiming near-parity with Anthropic and OpenAI, following Moonshot's Kimi K3, and that pace alone tells you the frontier is getting crowded faster than the benchmarks can be verified. Researchers should watch whether third-party evals on Qwen3.8-Max hold up once it's out in the wild, and whether "widely available" means real API access and pricing, not just a press release. Availability, not claims, is the test that counts.
Related Reading
Research & Benchmarks Google's FACTS benchmark shows 70% factuality ceiling across four tests
Research & Benchmarks Databricks finds multi-step agents beat single-turn RAG by 21% to 38% on STaRK
Research & Benchmarks Nvidia's DLSS 4.5 beta adds 6x Multi Frame Generation for RTX 50 GPUs
From Archives GPT-5.6 Helps Two Teams Solve Quantum Crypto Puzzle Within Hours
From Archives Alibaba's Qwen3.8-Max Writes 7,600 Lines of Code in Five Days
Common Questions Answered
What are the specifications of Alibaba's new Qwen3.8-Max model?
Qwen3.8-Max is Alibaba's largest and most capable AI model to date, featuring 2.4 trillion parameters. This scale places it in direct competition with top-tier systems from Anthropic, OpenAI, and Moonshot AI, making it one of the most advanced models currently available.
How does Alibaba claim Qwen3.8-Max compares to Claude Fable 5?
Alibaba described Qwen3.8-Max as 'second only' to Anthropic's flagship Claude Fable 5 during its preview last month. The company has positioned the model as a direct competitor to one of the leading AI systems in the industry.
What concerns are raised about Alibaba's benchmark testing for Qwen3.8-Max?
The article highlights that Alibaba is grading its own homework by using internal benchmarks and crowdsourced leaderboards like Arena.AI rather than undergoing independent, adversarial testing. This raises questions about whether the model's claimed performance will hold up when developers use it in real-world production scenarios with unexpected edge cases.
Why does Alibaba's release of Qwen3.8-Max matter geopolitically?
The open release of another highly capable Chinese AI model intensifies tensions in Silicon Valley and Washington regarding AI safety management and the United States' technological edge over China. This development adds to ongoing concerns about how to safely manage advanced AI systems while maintaining competitive advantage.
Further Reading
- Alibaba Drops Another China AI Model With Breakthrough Performance - Bloomberg
- Alibaba's Qwen Unveils Preview of Flagship AI Model - Bloomberg
- Alibaba Says New AI Model Is Just Second to Anthropic's 'Fable 5' - The Wall Street Journal
- Alibaba's new AI model scores higher than OpenAI, Google rivals in coding ranking - South China Morning Post
- Alibaba unveils new Qwen3.5 model for 'agentic AI era' - Reuters
