Do You Use ChatGPT For Work? This Study Says It Gets Things Wrong 30% Of The Time

A new study ranks popular AI chatbots by reliability, and ChatGPT, despite its massive popularity, didn't even make the top five. The winner might shock you.

A new AI reliability study is making waves.

Source : ABPLIVE AI

Show Quick Read

Key points generated by AI, verified by newsroom

• Perplexity AI leads AI chatbot reliability, scoring 85/100.
• It boasts lowest hallucination rate at 13% and 100% uptime.
• Grok and DeepSeek follow, outperforming popular ChatGPT and Gemini.

With AI chatbots becoming a regular part of work life, a new study has taken a closer look at which ones actually hold up when it matters. The findings, from US-based digital marketing agency Legal Guardian Digital, rank popular AI assistants based on how reliable they are for everyday tasks. The results may surprise many users: Perplexity AI came out on top, ahead of more widely recognised names like ChatGPT, Google Gemini, and Claude. 

The study assessed chatbots on hallucination rates, customer satisfaction, response consistency, and uptime reliability.

How Did Each Chatbot Perform In The Reliability Rankings?

According to the study, Perplexity AI recorded the lowest hallucination rate among major chatbots at 13%, well below the industry average of 22%. The platform also maintained a 100% uptime rate during the testing period, meaning it did not go down once. 

Elon Musk's Grok ranked second, with a hallucination rate of 15% and a perfect uptime score as well. Chinese AI chatbot DeepSeek came in third, with a 14% hallucination rate and a 99.52% uptime rate.

Perplexity AI achieved the highest overall reliability score of 85 out of 100. Grok followed at 79, and DeepSeek came in at 76.

What About ChatGPT, Gemini, And The Rest?

Despite being the most widely used AI chatbot globally, ChatGPT ranked sixth in the reliability index. The report found that ChatGPT generated incorrect responses in around 30% of cases, roughly double the error rate recorded by DeepSeek. 

Even so, it scored 4.7 out of 5 in customer ratings. Google Gemini ranked eighth, Meta AI came in ninth, and Claude, developed by Anthropic, ranked seventh, with comparatively more outages than its competitors. ChatGPT scored 50 out of 100 overall, while Gemini received 41.

On pricing, Perplexity AI reportedly costs $40 per month, and Grok costs $30, while DeepSeek is free to use despite its strong reliability ranking.

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 platforms such as Perplexity and Grok performed better.

OpenAI Launched A New Tool To Detect AI Images: We Tested It & Here's What Happened