DeepSeek Rations Its App for Lack of Compute: Daily Quotas, a 40,000-Character Cap and No More Week-Old Chats

DeepSeek is rationing its consumer app. According to Jiemian, citing STAR Market Daily (科创板日报), users reported on 16 November that the DeepSeek app and web chat now show a notice saying that, because of insufficient computing power, long conversations inactive for more than seven days are no longer supported.

Users also report two other changes. The maximum length of a single input has been cut to about 40,000 Chinese characters, and there is now a daily usage quota. When it runs out, the app says that “because the server is busy, the usage quota has been reached”. DeepSeek hasn’t made a formal announcement, so these are user reports carried by Chinese tech press, not an official policy page.

Rationing the app while the API grows

The timing is what makes this notable. In the same week, DeepSeek has been expanding its developer offering hard:

So the free consumer chat gets less compute, while paid API traffic and search calls get more. That’s a commercial choice dressed up as a capacity problem. GPUs are scarce in China, and DeepSeek is putting them where the money is.

Why it matters for search and SEO

DeepSeek’s app has been one of China’s biggest AI “search” surfaces. People ask it questions they used to type into Baidu. Every limit on the consumer product pushes those users somewhere else. The obvious winners are Tencent’s Yuanbao (which just made its search product official), ByteDance’s Doubao and Alibaba’s Qwen app. Alibaba launched Qwen4-Max on the same day and claims the average cost per task is down 60%.

At the same time, DeepSeek’s search index is likely to reach more products, not fewer. Cheap search and crawl APIs mean thousands of Chinese apps and agents may ground their answers in DeepSeek’s index without ever showing a DeepSeek logo. Tech Times reports that DeepSeek’s search scores results for relevance, authority and freshness, and claims to filter SEO spam and GEO manipulation. The index also hasn’t published a crawler user agent.

My take

Don’t read this as DeepSeek fading. Read it as DeepSeek changing what it is: less a consumer chatbot, more search and model infrastructure. For brands targeting China that means two things:

  • Track AI visibility across several assistants. Yuanbao, Doubao, Qwen and ERNIE Assistant are all gaining as DeepSeek’s app tightens. A brand audit that only checks DeepSeek’s chat is now incomplete.
  • Treat DeepSeek’s index as infrastructure. Make sure your Chinese-language pages are crawlable, clearly structured and authoritative, because the “authority tier” DeepSeek assigns to your host may decide whether hundreds of downstream apps cite you.

If the quotas become permanent, expect DeepSeek to launch a paid consumer tier soon. Either way, it’s a sign of how tight compute in China has become.

Dan Taylor is the founder and editor of International Search Insights (ISI), the publication behind ReadIsi, and a Partner and Head of Innovation (Search & AI) at SALT.agency. He has worked in search for more than a decade, in-house and agency-side, and is regarded as one of the West’s leading voices on search engines beyond Google, including Yandex, Baidu, Naver and Cốc Cốc. On ReadIsi, Dan covers the engines and AI answer platforms that English-language coverage tends to miss: Baidu and ERNIE, Yandex and Alice AI, Naver, Cốc Cốc, Shenma, Sogou, WeChat and Weibo search, plus DeepSeek and Alibaba’s Qwen. He also keeps dated records of documented changes, such as the Yandex, Baidu and Naver algorithm update histories and the DeepSeek, ERNIE and Qwen model update histories. His approach is to take learnings from global search engines and apply logic with a high evidence threshold. A claim needs a source, patterns seen in one market are treated as a hypothesis until they are checked in another, and where the evidence is thin he says so. For brands, that means advice built on what can be shown, with measurement that separates being cited from being chosen. Dan is credited with coining the term “edge SEO” and won the inaugural TechSEO Boost research competition in 2018 for his work on CDN and edge changes for SEO. In 2026 he was credited with coining “Infinite Tail” to describe how keyword research is changing with AI search. He is the author of AI Search Marketing For The Travel Sector, writes for Search Engine Journal, Search Engine Land and Travel Marketing Guy, and has advised technology, SaaS, travel and enterprise ecommerce brands, including Cloudflare, GitLab and Airbnb. He has given multiple talks and panels at the International Search Summit in Barcelona, and has also spoken at Optimization Moscow, brightonSEO, Google Search Central Live Deep Dive Asia in Bangkok, Athens SEO and the Asia SEO Virtual Summit 2026, where his talk looked at how Baidu, Naver, DeepSeek, Tencent, Cốc Cốc and Yandex built their own AI answers.