Recently, the Chinese AI startup DeepSeek @deepseek_ai has garnered significant attention, from its launch to its performance rivaling GPT, showcasing the strength of the Chinese team in the AI world.
The journey of DeepSeek began with the release of DeepSeek Coder in November 2023, an open-source model designed specifically for coding tasks. This was followed by DeepSeek LLM, a 67B parameter model aimed at competing with other large language models. DeepSeek-V2, launched in May 2024, drew attention for its strong performance and low cost, triggering a price war in the Chinese AI model market. This disruptive pricing strategy forced other major Chinese tech giants like ByteDance, Tencent, Baidu, and Alibaba to lower their AI model prices to remain competitive.
The successor to DeepSeek-V2 is DeepSeek-Coder-V2, a more advanced model with 236 billion parameters. It is designed to tackle complex coding challenges with a context length of up to 128K tokens. The model is available through a cost-effective API, priced at $0.14 per million input tokens and $0.28 per million output tokens.
The company's latest models, DeepSeek-V3 and DeepSeek-R1, further solidify its position as a disruptive force. DeepSeek-V3 is a 671B parameter model that performs exceptionally well across various benchmarks while requiring significantly fewer resources than its peers. DeepSeek-R1, released in January 2025, focuses on reasoning tasks and challenges OpenAI's o1 model with its advanced capabilities.
DeepSeek also offers a series of streamlined models called DeepSeek-R1-Distill, which are based on popular open-weight models like Llama and Qwen, and are fine-tuned on synthetic data generated by R1. These streamlined models provide different levels of performance and efficiency to meet various computational needs and hardware configurations.
Back to the blockchain field, the explosive popularity of AI Agents has spawned many meme projects with valuations exceeding one billion. As a new generation of more economical LLM, will DeepSeek suppress or even reshape the current AI landscape?
AI Agent is an intelligent system capable of autonomously executing tasks and interacting with the environment. LMM can serve as one of the core components of the Agent, providing powerful language processing capabilities that enable better understanding and generation of human language, thus playing a role in dialogue, recommendation, analysis, and other scenarios.
Agent relies on LLM to understand and generate natural language. LLM provides powerful language processing capabilities, allowing the Agent to interact with users in natural language, understand user needs, and generate corresponding responses. It can be said that LMM is an important technical support for AI Agents to achieve intelligent language interaction. From this perspective, the operation of Agents requires the support of LLM, but the quality of their expression and interaction does not solely depend on LLM.
Although DeepSeek's recent explosion has caused a certain decline in the AI sector, in the long run, breakthroughs in any sector have a huge impact on the collective progress of technology. The rapidly iterating AI Agents have already integrated swiftly with DeepSeek's updates. It can be said that AI projects that prioritize the use of DeepSeek will reap the benefits first; the new model will not disrupt the Agents but rather accelerate their development.
Currently, the brief decline in the AI sector will lead to accelerated development after fully absorbing DeepSeek. Shaw put it well: 'More powerful models are always good for Agents. For years, major AI laboratories have been surpassing each other. Sometimes Google leads, sometimes it's OpenAI, sometimes Claude, and today it's DeepSeek...'
#CryptoNewss #DeepSeek #ai16z