Just over two weeks after Google dropped their 3.1 update, OpenAI couldn't sit still and hastily pushed their GPT-5.4 model to the table. According to the latest benchmark tests, 5.4 is a significant leap over the previous 5.3 generation. Notably, its "Thinking" mode edges out Google's Gemini 3.1 Pro. And if you are paying for the GPT Pro subscription, you get access to 5.4 Pro, which boasts even better performance than the standard Thinking version. It looks like OpenAI has temporarily reclaimed the AI throne.
(Pictured: OpenAI's official benchmark results)
The AI "Turn-Based" Game: Honestly, Claude and Google are probably scrambling right now to prep their next counter-attack. Lately, the AI industry feels like an endless stream of minor, incremental updates—like squeezing a tube of toothpaste. We are all waiting for a massive, game-changing leap. Instead, it's just Google throwing a punch at OpenAI, OpenAI taking a step forward, and then accidentally stepping on Claude's toes in the process (just kidding, but you get the point!).
Comments NOTHING