Skip to content

AI Model Trained for Just $5.6M Outperforms Giants

A new AI model, DeepSeek V3, has emerged as a formidable competitor in the field of data science. Developed by a Chinese startup, DeepSeek, this model matches the performance of industry leaders like ChatGPT, Llama, and Claude but at a significantly lower cost. While OpenAI spent over $100 million to train its GPT-4 model, DeepSeek V3 was trained for just $5.6 million. This cost efficiency extends to API pricing, where DeepSeek’s chatbot model (V3) costs $0.14 per million tokens, and its reasoning model (R1) costs $0.55 per million tokens. In comparison, OpenAI’s gpt-4o API is priced at $2.50 per million input tokens, and the o1 API at $15.00 per million input tokens. The model’s capabilities were tested against established benchmarks in SQL queries, Exploratory Data Analysis (EDA), and Machine Learning (ML), showing promising results.

Source: medium.com

Related Links

Related Videos