According to the latest insights from tech expert Bindu Reddy on X, open-source artificial intelligence models are undergoing exponential growth. Notably, this segment's token consumption is rapidly narrowing the gap with Google's Gemini model series.
Developments
In her post, Reddy highlighted several prominent models leading this trend, including Kimi, DeepSeek, and GLM. These standout models are recognized for their excellent performance, capable of effectively handling around 50% of current routine tasks.
The token consumption of open models closing in on Gemini versions signals a major shift among developers. Rather than relying entirely on closed proprietary systems, the community is prioritizing operational cost optimization through more flexible alternatives.
Why This Matters
For the AI development community in Vietnam, the rise of models like DeepSeek or GLM opens up significant opportunities to optimize infrastructure costs. The ability of an open model to efficiently handle half of common tasks allows small and medium-sized enterprises to easily integrate AI into their workflows without worrying about hefty licensing fees from tech giants.