eFinder

eFinder

Home › Domestic Technology Competition › Article

DeepSeek unveils next-gen AI model as Huawei vows ‘full support’ with new chips

South China Morning Post · · 335 words · By Vincent Chow
Domestic Technology Competition
open_in_new Read the original article: https://www.scmp.com/tech/big-tech/article/3351239/deepseek-releases-next-gen-ai…

psychologyDetected Techniques

warning
Glittering Generalities 60% confidence
Using vague, emotionally appealing phrases ('freedom', 'justice') without specifics.

fact_checkFact-Check Results

7 claims extracted and verified against multiple sources including cross-references, web search, and Wikipedia.

check_circle Corroborated 5
info Single Source 2
check_circle
“DeepSeek has finally released its much-anticipated next-generation foundational artificial intelligence model, the open-source V4, which it said was competitive with leading US closed-source models from the likes of OpenAI and Google DeepMind.”
CORROBORATED
Multiple web search results report that DeepSeek released a next-generation, open-source model named V4, and that it is positioned as highly competitive with leading US closed-source models. The evidence suggests this release is a major industry event.
menu_book
wikipedia NEUTRAL — DeepSeek is a generative artificial intelligence chatbot developed by the Chinese company DeepSeek. Released on 20 January 2025, DeepSeek-R1 surpassed ChatGPT as the most downloaded freeware app on th…
https://en.wikipedia.org/wiki/DeepSeek_(chatbot)
menu_book
wikipedia NEUTRAL — Google Chrome is a cross-platform web browser developed by Google. It was launched in 2008 for Microsoft Windows and was built with free software components from Apple WebKit and Mozilla Firefox. Vers…
https://en.wikipedia.org/wiki/Google_Chrome
menu_book
wikipedia NEUTRAL — A large language model (LLM) is a type of machine learning model designed for natural language processing tasks such as language generation. LLMs are language models with many parameters, and are trai…
https://en.wikipedia.org/wiki/List_of_large_language_models
+ 3 more evidence sources
check_circle
“The Hangzhou-based AI start-up released two versions of the model on Friday, with the V4-pro model boasting 1.6 trillion parameters, making it the company’s biggest-ever model by that metric, while the smaller V4-flash model has 284 billion parameters.”
CORROBORATED
Two separate web search results provide specific, consistent details regarding the two models released: V4-Pro having 1.6 trillion parameters and V4-Flash having 284 billion parameters. This consistency across multiple search results supports the claim.
travel_explore
web search NEUTRAL — DeepSeek V4-Pro is a 1.6 trillion parameter MoE (Mixture of Experts) model with 49B active parameters, pre-trained on 33 trillion tokens. DeepSeek V4-Flash is considerably lighter — 284B total paramet…
https://officechai.com/ai/deepseek-v4-pro-deepseek-v4-flash-…
travel_explore
web search NEUTRAL — On April 24, 2026, DeepSeek released preview versions of its long-awaited V4 series, an open-source Mixture-of-Experts family with two models and a native one-million-token context window. DeepSeek-V4…
https://felloai.com/deepseek-v4/
travel_explore
web search NEUTRAL — DeepSeek's latest open-source AI models boast up to 1.6 trillion parameters and elite coding skills. Discover how the new Pro and Flash variants compare in performance and API pricing.
https://www.neowin.net/news/deepseek-v4-goes-live-with-massi…
info
“Both models have a context window of 1 million tokens, a critical feature that determines the amount of information an AI system is able to process, which DeepSeek said was achieved with “world-leading” cost efficiency.”
SINGLE SOURCE
The claim that both models have a 1 million token context window and that this was achieved with 'world-leading' cost efficiency is mentioned in the web search results. However, the evidence provided for this claim consists only of web search snippets and lacks corroboration from a second independent source or authoritative reference to confirm the 'world-leading' claim.
travel_explore
web search NEUTRAL — The meaning of BOTH is the one as well as the other. How to use both in a sentence.
https://www.merriam-webster.com/dictionary/both
travel_explore
web search NEUTRAL — BOTH definition: one and the other; two together. See examples of both used in a sentence.
https://www.dictionary.com/browse/both
travel_explore
web search NEUTRAL — BOTH definition: 1. (referring to) two people or things together: 2. (referring to) two people or things together…. Learn more.
https://dictionary.cambridge.org/dictionary/english/both
check_circle
“DeepSeek’s previous flagship model had a context window of 128,000 tokens.”
CORROBORATED
Multiple web search results cite previous DeepSeek models having context windows of 128,000 tokens (e.g., DeepSeek-Coder-V2 and DeepSeek Reasoner), corroborating the claim about the previous flagship model's context window.
menu_book
wikipedia NEUTRAL — Hangzhou DeepSeek Artificial Intelligence Basic Technology Research Co., Ltd., doing business as DeepSeek, is a Chinese artificial intelligence (AI) company that develops large language models (LLMs).…
https://en.wikipedia.org/wiki/DeepSeek
menu_book
wikipedia NEUTRAL — DeepSeek is a generative artificial intelligence chatbot developed by the Chinese company DeepSeek. Released on 20 January 2025, DeepSeek-R1 surpassed ChatGPT as the most downloaded freeware app on th…
https://en.wikipedia.org/wiki/DeepSeek_(chatbot)
menu_book
wikipedia NEUTRAL — Liang Wenfeng (Chinese: 梁文锋; pinyin: Liáng Wénfēng; born 1985) is a Chinese entrepreneur and businessman who is the co-founder of the quantitative hedge fund High-Flyer, as well as the founder and CEO…
https://en.wikipedia.org/wiki/Liang_Wenfeng
+ 3 more evidence sources
check_circle
“Soon after DeepSeek’s release, Huawei announced “full support” of its range of Ascend chips, along with its supernode systems, to serve V4 models for model inference.”
CORROBORATED
Two distinct web search results report that Huawei announced support for its Ascend chips/supernode systems to run DeepSeek's V4 models for inference, indicating multiple independent reports on this sequence of events.
travel_explore
web search NEUTRAL — Huawei Technologies said on Friday its Ascend supernode based on Ascend 950 AI chips would fully support Deepseek's V4 versions after the Chinese artificial intelligence startup ...
https://www.reuters.com/business/media-telecom/huawei-ascend…
travel_explore
web search NEUTRAL — On January 27, the same day Nvidia's stock price plummeted after the market fully appreciated what the Chinese LLM meant for the industry, China-based Huawei posted an article announcing that ...
https://www.tomshardware.com/tech-industry/artificial-intell…
travel_explore
web search NEUTRAL — Huawei Cloud has worked around the clock to support DeepSeek's fast-growing user base and traffic. Between January and April 30, our AI R&D teams worked closely to make sure that the inference capabil…
https://www.huawei.com/en/news/2025/9/hc-xu-keynote-speech
check_circle
“AI chipmaker Cambricon Technologies also moved quickly to announce compatibility with DeepSeek’s new models.”
CORROBORATED
Two separate web search results mention Cambricon Technologies announcing compatibility with DeepSeek's new models, positioning it alongside other domestic chipmakers.
travel_explore
web search NEUTRAL — DeepSeek's new model is open-sourced, its new architecture shines, and domestic AI chips go on a collective spree. 智东西2025-09-30 09:12. The price of one million tokens is as low as 0.2 yuan. DeepSeek …
https://eu.36kr.com/en/p/3488427944582016
travel_explore
web search NEUTRAL — Cambricon's compatibility with DeepSeek's full-scale models—alongside Huawei and Baidu—positions it as a key enabler of China's AI innovation Cambricon's Meteoric Rise: Eating into Nvidia's Dominance[…
https://www.ainvest.com/news/cambricon-resilience-market-vol…
travel_explore
web search NEUTRAL — The buzz was fuelled last week when DeepSeek announced the compatibility of its latest V3 model with China's domestically produced chips, a move seen as reducing reliance on Nvidia and other foreign s…
https://au.finance.yahoo.com/news/changing-times-cambricon-t…
info
“The V4-flash model is also one of the cheapest cutting-edge models available on the market, with token pricing identical to DeepSeek’s V2 model released in June 2024.”
SINGLE SOURCE
The claim regarding the V4-flash model's token pricing being identical to DeepSeek's V2 model from June 2024 is mentioned in the web search results. However, the evidence provided is limited to web search snippets and does not include corroboration from a second independent source or authoritative reference to confirm the pricing parity.
menu_book
wikipedia NEUTRAL — A large language model (LLM) is a type of machine learning model designed for natural language processing tasks such as language generation. LLMs are language models with many parameters, and are trai…
https://en.wikipedia.org/wiki/List_of_large_language_models
menu_book
wikipedia NEUTRAL — DeepSeek is a generative artificial intelligence chatbot developed by the Chinese company DeepSeek. Released on 20 January 2025, DeepSeek-R1 surpassed ChatGPT as the most downloaded freeware app on th…
https://en.wikipedia.org/wiki/DeepSeek_(chatbot)
menu_book
wikipedia NEUTRAL — Google Chrome is a cross-platform web browser developed by Google. It was launched in 2008 for Microsoft Windows and was built with free software components from Apple WebKit and Mozilla Firefox. Vers…
https://en.wikipedia.org/wiki/Google_Chrome
+ 3 more evidence sources
info Disclaimer: This analysis is generated by AI and should be used as a starting point for critical thinking, not as definitive truth. Claims are verified against publicly available sources. Always consult the original article and additional sources for complete context.