DeepSeek's V4-Flash AI model is cheapest to run among known models, research says
window 24h
IN SHORT
Research indicates that DeepSeek's V4-Flash AI model offers the most cost-effective performance among leading AI models, achieving a score of 50 on the Artificial Intelligence Index. Meanwhile, questions arise regarding Alibaba's Qwen3.8 Max model due to a lack of independent benchmarks to verify its claimed capabilities. Separately, a report suggests a potential U.S. ban on Chinese AI models could cost American businesses $12 billion annually by disrupting their reliance on these technologies.
✉Newsletter
PiQ Daily
Pick your topics. Get only what matters, on your cadence.
Key Numbers
50score on Artificial Analysis Intelligence Index
100maximum score on Artificial Analysis Intelligence Index
$12Bannual losses for U.S. businesses from AI ban
Who's Involved
DeepSeek
Chinese startup developing the V4-Flash AI model
Artificial Analysis
research firm that evaluated AI models
Alibaba
company that developed the Qwen3.8 Max AI model
Anthropic
company with AI models Claude Fable 5 and Claude
OpenAI
company with AI model GPT-5.6 Sol
1 / 2
Key facts
DeepSeek's V4-Flash AI model is the cheapest to run among known models.
DeepSeek's V4-Flash AI model scored 50 out of 100 on the Artificial Intelligence Index.
Alibaba's Qwen3.8 Max AI model claims to match unsupervised working skills of rivals.
Independent benchmarks for Alibaba's Qwen3.8 Max model are lacking.
A potential U.S. ban on Chinese AI models could cost businesses $12 billion annually.
The ban would impact companies relying on Chinese AI technologies.
Anthropic's Claude Fable 5 and OpenAI's GPT-5.6 Sol are mentioned as competitors.
The research was conducted by Artificial Analysis.
Research from Artificial Analysis suggests that DeepSeek's V4-Flash AI model is the most economical to operate compared to other known models. This Chinese startup's model achieved a score of 50 out of 100 on the firm's Intelligence Index, positioning it as a cheaper alternative to competitors such as Anthropic's Claude Fable 5 and OpenAI's GPT-5.6 Sol.
In parallel, Alibaba's new Qwen3.8 Max AI model faces scrutiny over its performance claims. While the model is said to match unsupervised working skills advertised by rivals like Anthropic's Claude, the absence of independent benchmarks has led to doubts about its actual capabilities.
Furthermore, a separate report highlights the potential economic impact of a U.S. ban on Chinese artificial intelligence models. Such a prohibition could lead to annual losses of $12 billion for American businesses that depend on these AI technologies for their operations. The ban would affect companies utilizing these models for various business functions.
↳ Why This Matters
Research from Artificial Analysis suggests that DeepSeek's V4-Flash AI model is the most economical to operate compared to other known models. This Chinese startup's model achieved a score of 50 out of 100 on the firm's Intelligence Index, positioning it as a cheaper alternative to competitors such as Anthropic's Claude Fable 5 and OpenAI's GPT-5.6 Sol.
Frequently asked questions
DeepSeek's V4-Flash is a new AI model released by the Chinese startup, designed to be ultra-low-cost to run.
According to Artificial Analysis, V4-Flash is significantly cheaper to run than models from Anthropic, OpenAI, and Moonshot AI, with an average cost of 3 cents per test.
The Intelligence Index is a benchmark developed by Artificial Analysis that combines results from nine tests covering coding, reasoning, and workplace assignments, scoring models out of 100.
The research included models such as Anthropic's Claude Fable 5 and Claude Opus 5, OpenAI's GPT-5.6 Sol, Moonshot AI's Kimi K3, Google's Gemini 3.6 Flash, Meta's Muse Spark 1.1, and Z.AI's GLM-5.2.
What Happens Next
01DeepSeek is preparing to release a more powerful version of its model, V4-Pro.
Get the newsletter.
Pick the topics you actually care about. We'll email when there's news worth your time, on the cadence you choose. Cancel any time from your account.
