Quick Take
  • Alibaba launched Qwen4 Max on Monday at roughly a third of its predecessor’s cost per task.
  • The cheaper model arrives as Alibaba’s Hong Kong shares are down 38.59% since January 1.
  • For holders of Alibaba stock, the lower price carries a trade-off.
  • Each task now earns Alibaba less, so total revenue falls unless usage rises enough to cover the gap.

What Happened

Alibaba launched Qwen4 Max on Monday at roughly a third of its predecessor’s cost per task. The cheaper model arrives as Alibaba’s Hong Kong shares are down 38.59% since January 1.

Hong Kong trading on Monday gave investors their first chance to price the launch.

What Does the Launch Mean for Alibaba Stock?

By comparison, the shares rose 6.15% by the same break on the day Qwen3.8 Max launched in August. The shares sit nearly 19% below HK$112.70, the price set for Alibaba’s $10 billion share sale in August.

Market Context

For holders of Alibaba stock, the lower price carries a trade-off. Each task now earns Alibaba less, so total revenue falls unless usage rises enough to cover the gap.

The cost gap with Qwen3.8 Max is wider than the score gap. Alibaba cut per-token prices by 25% and cached-input prices by 70%.

“Qwen4 Max uses ~105k output tokens per task, 39% fewer than Qwen3.8 Max (~171k), and 32% fewer input tokens to run the Index (9.9B vs 14.5B). Input makes up ~85-90% of the cost of running the Index for both models since agentic evaluations re-send context on every turn, so the 70% lower cache hit price ($0.075 vs $0.25 per 1M tokens) and 25% lower input price ($1.50 vs $2.00) for Qwen4 Max reduce Cost per Task,” Artificial Analysis explained.

Those efficiency gains could soften the margin hit from lower prices. Still, it is too early to tell whether usage will grow enough to offset the lower prices.

However, the period covers Qwen3.8 Max’s first two months on sale. Any MaaS update would indicate how model demand tracked before the latest price cut.

Why It Matters

These cost figures matter to shareholders because Alibaba expects selling AI models to become its biggest cloud revenue source.

Details

How Much Smarter and Cheaper Is Qwen4 Max?

Artificial Analysis, an independent AI benchmarking firm, scored Qwen4 Max at 43 on its Intelligence Index. That is nine points above Qwen3.8 Max.

On Arena’s Agent leaderboard, built on real user tasks, Qwen4 Max ranks 14th and first among Chinese models.

Together with lower token use, the total savings vary by test.

On the Artificial Analysis index, cost per task fell 67%, from $10.02 to $3.27.

On Work360, Alibaba’s own test of industry tasks, the average cost per task fell 62%, from $4.43 to $1.67.

Alibaba says token use fell by 24% to 35% on Work360, depending on the reasoning setting. On Terminal-Bench 4.0, a test of command-line tasks, it fell by about 48% to 49%.

Follow us on X to get the latest news as it happens

Can a Cheaper Qwen Lift Alibaba Cloud Revenue?

Alibaba sells Qwen through Alibaba Cloud as model-as-a-service (MaaS), with per-token pricing for API customers. Its model and application services, including MaaS, passed RMB 16 billion in annual recurring revenue by August.

Qwen4 Max activates 25% fewer parameters per token than Qwen3.8 Max. Alibaba says its new design improves computing efficiency.

Alibaba has also put the model to work on its own cloud. In one case, the model cut the cost of scheduling model calls for data labeling by 10.42% on average.

Alibaba’s Hong Kong shares rose 1.05% to HK$91.50 by Monday’s lunch break, from Friday’s HK$90.55 close.

Alibaba reports September-quarter results on Thursday, November 19. That quarter ended before Qwen4 Max arrived, so the new model is absent from the numbers. The first results to include it will cover the December quarter.