<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>DEV Community: Bison</title>
    <description>The latest articles on DEV Community by Bison (@askbison).</description>
    <link>https://dev.to/askbison</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4254314%2Fc2aeec10-9878-4c09-800e-2af42b4812fc.png</url>
      <title>DEV Community: Bison</title>
      <link>https://dev.to/askbison</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/askbison"/>
    <language>en</language>
    <item>
      <title>Send one prompt to GPT, Claude, Gemini and Grok with 40 lines of Node (and the reasoning-token trap I hit)</title>
      <dc:creator>Bison</dc:creator>
      <pubDate>Thu, 19 Nov 2026 18:15:09 +0000</pubDate>
      <link>https://dev.to/askbison/send-one-prompt-to-gpt-claude-gemini-and-grok-with-40-lines-of-node-and-the-reasoning-token-trap-20ij</link>
      <guid>https://dev.to/askbison/send-one-prompt-to-gpt-claude-gemini-and-grok-with-40-lines-of-node-and-the-reasoning-token-trap-20ij</guid>
      <description>&lt;p&gt;I build &lt;a href="https://askbison.com/r/devto2" rel="noopener noreferrer"&gt;Bison&lt;/a&gt;, a chat app that compares AI models side by side. Today I wrote the small script behind our public comparison pages, and it is useful on its own: one prompt, four models, all answers in parallel. This is the core of it, tested with Node 20+ and &lt;a href="https://openrouter.ai" rel="noopener noreferrer"&gt;OpenRouter&lt;/a&gt; (one API key, many providers).&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight javascript"&gt;&lt;code&gt;&lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="no"&gt;KEY&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="nb"&gt;process&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="n"&gt;env&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="no"&gt;OPENROUTER_API_KEY&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt;
&lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="no"&gt;MODELS&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;
  &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="n"&gt;id&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"openai/gpt-6.2-sol"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;name&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"GPT"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="n"&gt;id&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"anthropic/claude-sonnet-5.5"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;name&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Claude"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="n"&gt;id&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"google/gemini-4-argon"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;name&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Gemini"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
  &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="n"&gt;id&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"x-ai/grok-4.7"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;name&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"Grok"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
&lt;span class="p"&gt;]&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt;

&lt;span class="k"&gt;async&lt;/span&gt; &lt;span class="kd"&gt;function&lt;/span&gt; &lt;span class="nf"&gt;ask&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;model&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;prompt&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
  &lt;span class="k"&gt;for&lt;/span&gt; &lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="kd"&gt;let&lt;/span&gt; &lt;span class="n"&gt;attempt&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="mi"&gt;0&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt; &lt;span class="n"&gt;attempt&lt;/span&gt; &lt;span class="o"&gt;&amp;lt;&lt;/span&gt; &lt;span class="mi"&gt;3&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt; &lt;span class="n"&gt;attempt&lt;/span&gt;&lt;span class="o"&gt;++&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
    &lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="n"&gt;res&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="k"&gt;await&lt;/span&gt; &lt;span class="nb"&gt;fetch&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="s2"&gt;"https://openrouter.ai/api/v1/chat/completions"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt;
      &lt;span class="n"&gt;method&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"POST"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
      &lt;span class="n"&gt;headers&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="nc"&gt;Authorization&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="sb"&gt;`Bearer ${KEY}`&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"Content-Type"&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"application/json"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
      &lt;span class="n"&gt;body&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="nb"&gt;JSON&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;stringify&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="p"&gt;{&lt;/span&gt;
        &lt;span class="n"&gt;model&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
        &lt;span class="n"&gt;max_tokens&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="mi"&gt;3000&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
        &lt;span class="n"&gt;reasoning&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="n"&gt;effort&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"low"&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
        &lt;span class="n"&gt;messages&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="p"&gt;{&lt;/span&gt; &lt;span class="n"&gt;role&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="s2"&gt;"user"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;content&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="n"&gt;prompt&lt;/span&gt; &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
      &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="p"&gt;}&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt;
    &lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="n"&gt;text&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="k"&gt;await&lt;/span&gt; &lt;span class="n"&gt;res&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;json&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="n"&gt;choices&lt;/span&gt;&lt;span class="o"&gt;?&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="mi"&gt;0&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;&lt;span class="o"&gt;?&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="n"&gt;message&lt;/span&gt;&lt;span class="o"&gt;?&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="n"&gt;content&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt;
    &lt;span class="k"&gt;if&lt;/span&gt; &lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;text&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="k"&gt;return&lt;/span&gt; &lt;span class="n"&gt;text&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;trim&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt;
    &lt;span class="k"&gt;await&lt;/span&gt; &lt;span class="k"&gt;new&lt;/span&gt; &lt;span class="nb"&gt;Promise&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;r&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="o"&gt;=&amp;gt;&lt;/span&gt; &lt;span class="nb"&gt;setTimeout&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;r&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="mi"&gt;1500&lt;/span&gt; &lt;span class="o"&gt;*&lt;/span&gt; &lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;attempt&lt;/span&gt; &lt;span class="o"&gt;+&lt;/span&gt; &lt;span class="mi"&gt;1&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt;
  &lt;span class="p"&gt;}&lt;/span&gt;
  &lt;span class="k"&gt;return&lt;/span&gt; &lt;span class="s2"&gt;"(no answer)"&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt;
&lt;span class="p"&gt;}&lt;/span&gt;

&lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="n"&gt;prompt&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="nb"&gt;process&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="n"&gt;argv&lt;/span&gt;&lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="mi"&gt;2&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt;
&lt;span class="kd"&gt;const&lt;/span&gt; &lt;span class="n"&gt;answers&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="k"&gt;await&lt;/span&gt; &lt;span class="nb"&gt;Promise&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;all&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="no"&gt;MODELS&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;map&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;m&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="o"&gt;=&amp;gt;&lt;/span&gt; &lt;span class="nf"&gt;ask&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;m&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="n"&gt;id&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;prompt&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt;
&lt;span class="no"&gt;MODELS&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;forEach&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;m&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;i&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="o"&gt;=&amp;gt;&lt;/span&gt; &lt;span class="nb"&gt;console&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;log&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="sb"&gt;`\n=== ${m.name} ===\n${answers[i]}`&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;;&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Run it with &lt;code&gt;OPENROUTER_API_KEY=... node compare.mjs "Explain inflation with a grocery example in under 100 words"&lt;/code&gt;. Check the current model IDs in the OpenRouter catalogue first, they change often.&lt;/p&gt;

&lt;h2&gt;
  
  
  The trap: reasoning models can return an empty answer
&lt;/h2&gt;

&lt;p&gt;My first version used &lt;code&gt;max_tokens: 600&lt;/code&gt;. One of the models kept coming back with &lt;code&gt;content: null&lt;/code&gt; and &lt;code&gt;finish_reason: "length"&lt;/code&gt;. The cause: models that think before answering spend part of the token budget on hidden reasoning, so a small limit gets used up before a single visible word is written. The fix was two lines: give a much bigger &lt;code&gt;max_tokens&lt;/code&gt; than the answer needs, and lower the reasoning effort (&lt;code&gt;reasoning: { effort: "low" }&lt;/code&gt;). Always handle &lt;code&gt;content === null&lt;/code&gt; instead of assuming a string.&lt;/p&gt;

&lt;p&gt;A second small one: some providers reject &lt;code&gt;temperature&lt;/code&gt; above 1, so keep it at 1 or below.&lt;/p&gt;

&lt;h2&gt;
  
  
  Why compare at all
&lt;/h2&gt;

&lt;p&gt;Single answers hide differences you only see side by side: in my small runs one model averaged about 116 words per answer and another about 181 for the same "under 170 words" instruction. I wrote up the longer version with full answers here: &lt;a href="https://dev.to/askbison/i-ran-the-same-10-prompts-through-gpt-claude-gemini-and-grok-here-is-the-pattern-254d"&gt;the 10-prompt comparison&lt;/a&gt;. If you would rather not run scripts, Bison does the same thing in a UI (free plan, no card), but the snippet above is all you need to start.&lt;/p&gt;

&lt;p&gt;What would you add: streaming, a judge model, cost tracking?&lt;/p&gt;

</description>
      <category>node</category>
      <category>ai</category>
      <category>llm</category>
      <category>tutorial</category>
    </item>
    <item>
      <title>I ran the same 10 prompts through GPT, Claude, Gemini and Grok. Here is the pattern.</title>
      <dc:creator>Bison</dc:creator>
      <pubDate>Thu, 19 Nov 2026 18:06:51 +0000</pubDate>
      <link>https://dev.to/askbison/i-ran-the-same-10-prompts-through-gpt-claude-gemini-and-grok-here-is-the-pattern-254d</link>
      <guid>https://dev.to/askbison/i-ran-the-same-10-prompts-through-gpt-claude-gemini-and-grok-here-is-the-pattern-254d</guid>
      <description>&lt;p&gt;I am the maker of &lt;a href="https://askbison.com/r/devto" rel="noopener noreferrer"&gt;Bison&lt;/a&gt;, a chat app that sends one prompt to several models. Because of that I read a lot of side-by-side answers, so I ran a small experiment: 10 everyday prompts (emails, explanations, a logic puzzle, a word problem, a content plan), the same wording sent once each to GPT-6.2 Sol, Claude Sonnet 5.5, Gemini 4 Argon and Grok 4.7, default settings, no cherry-picking.&lt;/p&gt;

&lt;p&gt;Full answers are public, so you can check my reading instead of trusting it:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;&lt;a href="https://askbison.com/arena/write-a-polite-follow-up-email-to-a-recruiter-i-haven-t-heard-from-in?ref=devto" rel="noopener noreferrer"&gt;Follow-Up Email After Interview&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://askbison.com/arena/explain-how-compound-interest-works-in-simple-terms-using-a-real-life?ref=devto" rel="noopener noreferrer"&gt;Explain compound interest simply&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://askbison.com/arena/explain-how-blockchain-works-in-simple-terms-using-an-everyday-analogy?ref=devto" rel="noopener noreferrer"&gt;Explain blockchain simply&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://askbison.com/arena/explain-how-to-solve-quadratic-equations-using-the-quadratic-formula-w?ref=devto" rel="noopener noreferrer"&gt;Explain the quadratic formula step by step&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://askbison.com/arena/solve-the-classic-river-crossing-puzzle-with-a-wolf-goat-and-cabbage-a?ref=devto" rel="noopener noreferrer"&gt;Wolf, Goat, Cabbage Puzzle Solved&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://askbison.com/arena/write-a-30-day-instagram-content-plan-for-my-new-handmade-candle-shop?ref=devto" rel="noopener noreferrer"&gt;Instagram content plan for a small shop&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://askbison.com/arena/two-trains-leave-cities-300-miles-apart-at-the-same-time-one-going-60?ref=devto" rel="noopener noreferrer"&gt;Solve the two trains word problem&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://askbison.com/arena/write-a-professional-email-to-my-manager-asking-for-a-one-week-extensi?ref=devto" rel="noopener noreferrer"&gt;Email asking for a deadline extension&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://askbison.com/arena/explain-how-inflation-works-in-simple-terms-using-a-grocery-shopping-e?ref=devto" rel="noopener noreferrer"&gt;Explain inflation simply&lt;/a&gt;&lt;/li&gt;
&lt;li&gt;&lt;a href="https://askbison.com/arena/help-me-write-a-clear-thesis-statement-for-a-5-paragraph-essay-arguing?ref=devto" rel="noopener noreferrer"&gt;Write a thesis statement for an essay&lt;/a&gt;&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  What I saw
&lt;/h2&gt;

&lt;p&gt;&lt;strong&gt;Length is the biggest tell.&lt;/strong&gt; I asked every model for "under 170 words". Average answer length across the 10 prompts: GPT 116 words, Grok 123, Gemini 159, Claude 181. Claude's average was the only one above the limit, and it was also the one adding extras: a verification step, a common-mistakes list, a sending tip.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;GPT is the "ready to send" model.&lt;/strong&gt; Tightest structure and the least to edit: on the deadline-extension email its version was the compact, confident, ready-to-send one.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Claude is the "teacher".&lt;/strong&gt; On the quadratic formula and the two-trains problem it added a check and an alternative method. Great for learning, a bit much if you just want the answer.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Gemini is the warm one.&lt;/strong&gt; Conversational tone, friendly analogies (the blockchain answer compared it to a shared Google Doc).&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Grok is the brief one.&lt;/strong&gt; Correct and short; in a few of the ten it used less of the detail from the prompt than the others (for example a generic subject line on the follow-up email).&lt;/p&gt;

&lt;h2&gt;
  
  
  The honest caveats
&lt;/h2&gt;

&lt;ul&gt;
&lt;li&gt;Ten prompts and one attempt each is an anecdote, not a benchmark. Re-run them and you will get different text.&lt;/li&gt;
&lt;li&gt;The written analysis on each page was produced by Claude, so it may favour its own style. Read the raw answers.&lt;/li&gt;
&lt;li&gt;The prompts were LLM-generated from topic lists, not collected from real users.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  Takeaway
&lt;/h2&gt;

&lt;p&gt;The practical lesson was not "model X wins". It was that the best model depends on whether you want brevity, a teaching answer, or a tone, and that you only notice when you see answers next to each other. If you want to try that yourself, &lt;a href="https://askbison.com/r/devto" rel="noopener noreferrer"&gt;Bison&lt;/a&gt; has a free plan, but the same experiment works with any tool that lets you paste one prompt into several models.&lt;/p&gt;

&lt;p&gt;What prompts would you want to see in the next round?&lt;/p&gt;

</description>
      <category>ai</category>
      <category>chatgpt</category>
      <category>llm</category>
      <category>productivity</category>
    </item>
  </channel>
</rss>
