News list for " Seek"
Today, we are releasing two official models simultaneously: DeepSeek-V3.2 and DeepSeek-V3.2-Speciale. DeepSeek-V3.2 is our first model that integrates thinking into tool use, and supports tool calls in both thinking and non-thinking modes. We propose a large-scale agent training data synthesis method, which constructs a large number of "difficult-to-answer, easy-to-verify" reinforcement learning tasks (1800 + environments, 85,000 + complex instructions), which greatly improves the generalization...
今天，我们同时发布两个正式版模型：DeepSeek-V3.2和DeepSeek-V3.2-Speciale。DeepSeek-V3.2是我们推出的首个将思考融入工具使用的模型，并且同时支持思考模式与非思考模式的工具调用。我们提出了一种大规模Agent训练数据合成方法，构造了大量“难解答，易验证”的强化学习任务（1800+环境，85,000+复杂指令），大幅提高了模型的泛化能力。（DeepSeek）
DeepSeek has launched a new mathematical reasoning model, DeepSeekMath-V2, using a self-validating training framework. The model is built on DeepSeek-V3.2-Exp-Base, automatically reviews the generated mathematical proofs through an LLM validator, and continuously optimizes performance using difficult samples. (Securities Times)
DeepSeek推出新型数学推理模型DeepSeekMath-V2，采用可自我验证的训练框架。该模型基于DeepSeek-V3.2-Exp-Base构建，通过LLM验证器自动审查生成的数学证明，并利用高难度样本持续优化性能。（证券时报）
The 2025 Forbes Chinese mainland rich list was officially released today, and two-thirds of the rich people on the list have achieved growth in their wealth. Among them, the founder of Nongfu Spring, Zhong Liangyi, has topped the list for the fifth consecutive year, and his wealth has increased by 26.30 billion US dollars to 77.10 billion US dollars (Note: the current exchange rate is about 548.965 billion yuan), which is the richest person in terms of net worth. Zhang Yiming, co-founder of Byte...
2025福布斯中国内地富豪榜今日正式发布，榜单中三分之二的富豪身家实现增长。其中，农夫山泉创始人钟睒睒连续第五年蝉联榜首，其财富增长263亿美元，达到771亿美元（注：现汇率约合5489.65亿元人民币），是身家金额增长最多的富豪。字节跳动联合创始人张一鸣身家增长237亿美元，以693亿美元（现汇...
nof1.ai founder Jay A posted on the X platform that the first season of Alpha Arena has officially ended, Qwen 3 MAX won the championship, with a balance of $12,252.72 in the account as of the press release, and a balance of $10,634.21 in DeepSeek, maintaining positive income. The GPT-5 balance of $3,534.53, Gemini 2.5 Pro balance of $3,980.75, Grok-4 balance of $5,758.49, and Claude Sonnet 4.5 balance of $7,294.59 were all negative returns.
nof1.ai创始人Jay A在X平台发文表示，Alpha Arena第一季已正式结束，Qwen 3 MAX获得冠军，截至发稿账户余额12,252.72美元，于此同时DeepSeek余额10,634.21美元，保持正收益。 而GPT-5余额3534.53美元、Gemini 2.5 Pro余额3980.75美元、Grok-4余额5758.49美元、Claude Sonnet 4.5余额7294.59美元均为负收益。
Nof1, an artificial intelligence research lab focused on financial marekt, began a large-scale model trading test AlphaArena on the 18th. The test uses six mainstream AI large models (GPT-5, Gemini 2.5 Pro, Grok-4, ClaudeSonnet 4.5, DeepSeek V3.1, Qwen3 Max), each of which receives $10,000 in real money on Hyperliquid, with the same hints and input data. As of press time, DeepSeek and Qwen3 Max have doubled their revenue and are leading the way.
