Grok Chatbot Tops AI Reliability Ranking: Only 8% 'Hallucinations'

AI Reliability Ranking

According to Главком: The Grok chatbot, developed by Elon Musk's xAI, topped the December 2025 AI reliability ranking. This chatbot demonstrated the lowest level of 'hallucinations' among analyzed neural networks, providing unreliable information only 8% of the time. This is significantly lower than the metrics of other popular AI systems.

For comparison, the error rates of other systems are as follows:

• ChatGPT by OpenAI - 35%
• Google Gemini - over 40%
• Microsoft Copilot - 27%
• Chinese neural network DeepSeek - 14%
• Search engine Perplexity - 13%
• Claude - 17%

This data underscores the importance of AI reliability in today's informational environment.

Trends in AI Usage

Research has shown that over 65% of companies in the US have already integrated AI into their workflows, indicating the growing popularity of these technologies. At the same time, according to an Apptopia analysis, the growth of the ChatGPT mobile application has slowed since April 2025, which may indicate shifts in consumer preferences.

In light of these trends, it is important to consider that while Grok demonstrates the highest reliability ratings, other systems still face serious challenges in providing accurate information. This issue remains pertinent for developers and users of artificial intelligence.

These results highlight the importance of developing AI technologies that can ensure high reliability and accuracy.

Competition among leading companies in this field spurs innovation and improves product quality. Given the increasing integration of AI into business processes, the reliability of information becomes critically important for the successful utilization of these technologies in various domains.