Alibaba and DeepSeek shake up AI race with new models at a fraction of Western costs

DeepSeek's V4-Flash model costs an average of just 3 cents per benchmark test, compared with $3.15 for Anthropic's Claude Fable 5, making it the world's cheapest well-known AI model to run, according to research firm Artificial Analysis. The finding underscores how Chinese AI firms are aggressively undercutting Western rivals on price while narrowing the performance gap.ndtvprofit+1

Rock-Bottom Pricing

DeepSeek charges $0.14 per million input tokens and $0.28 per million output tokens for V4-Flash, which moved into public beta on July 31. Artificial Analysis estimated the model's cost-per-task at roughly 60 times less than OpenAI's GPT-5.6 Sol, priced at $1.86 per test, and more than 100 times cheaper than Claude Fable 5. The benchmark-based comparison accounts for how much data a model must process and generate to complete a task, offering a more realistic measure of value than headline token pricing alone.dailysabah+3

V4-Flash scored 50 out of 100 on Artificial Analysis's Intelligence Index, matching Google's Alphabet Inc. ↗ 4.44% Gemini 3.6 Flash and trailing Meta's ↗ 6.02% Muse Spark 1.1 by a single point. The model is a 284-billion-parameter mixture-of-experts architecture with only 13 billion parameters active per token, enabling its low operating costs.marktechpost+1

Alibaba Joins the Race

On the same day the Artificial Analysis findings were published, Alibaba ↗ 4.13% unveiled Qwen3.8-Max, its largest AI model to date, with 2.4 trillion parameters and a one-million-token context window. The model climbed to the top-ranked Chinese position on the Arena.AI text leaderboard and second globally on its vision leaderboard, trailing only Anthropic's Fable 5 variants. Alibaba's Hong Kong-listed shares rose 7% on the news, according to Reuters.cnbc+1

Intensifying Competition

DeepSeek, which Reuters reported is preparing for a potential initial public offering, faces growing domestic competition from ByteDance, Moonshot AI, MiniMax, and Z.AI. Lian Jye Su, chief analyst at research firm Omdia, told Reuters that Chinese AI companies have identified a market for models that are "good enough, affordable, transparent and accessible," rather than the industry's absolute best. DeepSeek is also developing V4-Pro, a more powerful successor, though no launch date has been announced.ndtvprofit+1