<p>Google has released Gemini 2.5 Pro — an experimental version of the most advanced model that mentally articulates its answer before delivering it. This is not just generation, but full-fledged reasoning: analysis, conclusions, context, logic.</p>
<p>As a result, the model has reached the top position in LMArena (a rating based on human preferences) and has outperformed competitors in mathematical, scientific, and coding benchmarks.</p>
<p>What’s particularly interesting:</p>
<ul>
<li>18.8% on Humanity’s Last Exam — an exam on frontier skills devised by hundreds of experts;</li>
<li>63.8% on SWE-Bench Verified — one of the main tests for “agent programming”; </li>
<li>context window — up to a million tokens already now, two million — coming soon.</li>
</ul>
<p>Gemini 2.5 can generate complex web applications, edit and transform code, and even create games from a single prompt. Moreover, it does this based on reasoning, rather than just guessing tokens.</p>
<p>It will soon be available on Vertex AI. For now, you can play in Google AI Studio or in Gemini Advanced.</p>
<p>Video: <a href="https://www.youtube.com/watch?v=RLCBSpgos6s">https://www.youtube.com/watch?v=RLCBSpgos6s</a></p>;
<p>#gemini #googledeepmind #ai #reasoning #coding</p>
<p><a href="https://t.me/+OvImEUmA7W5mYTRi">————————— Мысли Рвачева —————————</a></p>
