3 min readNew DelhiMay 27, 2026 08:04 AM IST AI chatbots are everywhere, but which one is the most reliable? That is a question likely on the minds of millions of users. US-based digital marketing agency Legal Guardian Digital has released a new study ranking the most reliable chatbot for everyday work tasks. The findings reveal that Perplexity outperformed its more widely used peers such as ChatGPT, Google Gemini, and Claude.
The report assessed popular AI chatbots based on their hallucination rates, customer satisfaction, response consistency, and uptime reliability. The study arrives at a time when AI assistants are increasingly being used in workspaces. The report claims that one in four American workers use AI tools regularly.
Based on the study, Perplexity AI reported the lowest hallucination rate among major chatbots at 13 per cent which is significantly lower than the industry average of 22 per cent. The platform also maintained a 100 per cent uptime rate, meaning the service reportedly did not experience outages during the testing period.
Meanwhile, Elon Musk’s Grok ranked second with a hallucination rate of 15 per cent and a perfect uptime score. Chinese AI chatbot DeepSeek came in third, with a 14 per cent hallucination rate and a 99.52 per cent uptime rate.
Perhaps the biggest surprise has been ChatGPT. Despite being the most widely used AI chatbot globally, it ranked sixth in the reliability index. The report claimed ChatGPT generated incorrect responses in around 30 per cent of cases, about double the error rate recorded by DeepSeek. Regardless, ChatGPT continued to score highly in customer ratings, obtaining 4.7 out of 5 from users.
On the other hand, Google Gemini ranked eighth on the list, while Meta AI came in ninth. The report noted that Claude, developed by Anthropic, ranked seventh and experienced comparatively more outages than its competitors.
The team behind the study claimed that they combined several factors to create an overall reliability score, including how often chatbots returned false information, their response consistency across different prompts, customer reviews, and service uptime.Story continues below this ad
Perplexity AI achieved the highest overall reliability score of 85 out of 100, followed by Grok at 79 and DeepSeek at 76. ChatGPT scored 50, while Google Gemini received 41. The report also highlighted costing differences between AI services. Perplexity AI reportedly costs $40 per month, while Grok costs $30. DeepSeek was described as free to use, despite its third ranking among chatbots in terms of reliability.
Austin Hunt, CEO of Legal Guardian Digital, said users often assume ChatGPT is the most reliable AI assistant because of its popularity and early market lead. However, he argued that when it came to reliability and uptime, smaller AI platforms such as Perplexity and Grok performed better.
Expand © IE Online Media Services Pvt Ltd