Is ChatGPT Losing Its Edge? Find Out What Our Readers Think!

As we recently discussed, OpenAI’s models have progressively lost ground in the LMArena rankings, a system that assesses AI performance based on user votes. By December 2025, the company’s top model ranked eighth. By January 2026, it had slipped to ninth place. The situation worsened by February: not a single OpenAI model was in the top 10, with GPT-5.1 high landing at eleventh.

Does this decline in rankings reflect a shared sentiment among users? To find out, we conducted two surveys: one with our readers at the end of this article, and another on LinkedIn. The results confirm a significant trend.

Perceived quality decline among users

On LinkedIn, we polled our community from January 27 to February 3, 2026. We asked, “Do you think ChatGPT is less effective than before?” Of the 267 respondents, 119 stated they had replaced the tool, which represents 45% of the participants. Only 51 voters, about 19%, believe that ChatGPT remains the best. The remaining 97 respondents, roughly 36%, have not compared it with other available solutions.

Similarly, we posed a related question to readers of BDM between January 27 and February 17, 2026. Of the 101 respondents, 50 noted a decline in ChatGPT’s performance over time. 42 did not observe any degradation, and 9 stated that they do not use the tool.

“In some aspects, ChatGPT remains the best. But for more editorial tasks or in developing certain codes, I find that Claude and Perplexity handle information processing more advancedly,” explains Kelly Favart on LinkedIn.

Excluding non-users, more than half of the respondents (54%) perceive a decline in quality. This sentiment resonates with the regular critiques since the release of GPT-5 and its variants, often considered below expectations.

“I find it less effective but still use it (…). ChatGPT remains a very effective tool, but it lacks flexibility. Its competitors are making significant progress,” notes Rodolphe Gilbert-Collet on LinkedIn.

Gemini and Claude, the favored alternatives

In response to these observations, which solutions are users turning to? We asked them as well. Among 89 respondents, 46% have replaced ChatGPT with Gemini (41 responses), Google’s assistant. Claude, developed by Anthropic, is in third place with 13 responses (15%), following the 29 users who continue to use ChatGPT (33%). Perplexity and Grok are also mentioned, albeit more marginally. It’s worth noting that many users are diversifying their tools based on tasks and needs.

In total, 64% of respondents have switched to an alternative. This shift reflects a very real market dynamic. Gemini now claims 750 million active monthly users, up from 650 million the previous quarter. Its market share has increased from 5.7% to 21.5% in a year, while ChatGPT’s share has dropped from 86.7% to 64.5% over the same period, according to AIBars.

On Anthropic’s side, growth is more pronounced in the professional sector. Claude is gradually becoming more prominent among developers and businesses, with, according to Computer News, eight of the top ten Fortune 10 companies among its clients.

A market no longer monopolized

These developments paint a very different picture from early 2025, when OpenAI dominated unchallenged. Anthropic’s recent fundraising might be the most striking symbol of this: $30 billion for a valuation of $380 billion, surpassing Netflix’s market capitalization. This represents a strong signal of investor confidence in alternatives to OpenAI.

Of course, these surveys do not claim to be exhaustive. The samples are modest and represent a savvy audience, readers of BDM and digital professionals, who often mix different AI tools to achieve their goals. However, they reflect a shift in usage and perceptions. The generative AI market is no longer a monopoly; it has become a competitive field where each player must now prove their worth, model by model.