Cheaper AI technology: what does it mean to marketers

According to the report of Stanford HAI's AI Index Report, the cost of using AI models GPT-3.5 has decreased from $ 20 to $ 0.07 for a million tokens. This was made possible by the significant reduction in the cost of AI that has occurred in the last two years.

🚀 According to the report, large language models (LLM) have become cheaper from 9 to 900 times a year, depending on the task. This changed Roi for AI in marketing, making tools that were previously too expensive, income even for medium -sized companies.

💡 It is important to note that the difference between the best AI models decreases. The difference between the first and tenth models decreased from 11.9% to only 5.4% over the past year.

• 📌 The fall in prices makes it possible to create and optimize large volume content.
• 📌 More AI models can handle up to 1-2 million tokens at a time, allowing you to analyze the goals of websites to gain competitive advantages.
• 📌 Improves knowledge management using Retrieval-Augmented Generation (RAG) when AI receives information from your company data.