https://backend-engineering-strategy-tools.github.io/site/tags/chatgpt/Recent content in Chatgpt on Backend Engineering Strategy ToolsHugo -- gohugo.ioen-usThu, 14 May 2026 00:00:00 +0000https://backend-engineering-strategy-tools.github.io/site/public-notes/ai/ai-chat/Thu, 14 May 2026 00:00:00 +0000https://backend-engineering-strategy-tools.github.io/site/public-notes/ai/ai-chat/<p>Browser-based chat interfaces to large language models. None of these require setup — open a tab and start typing. The differences are in what each is good at, not in how they work.</p> <hr> <h2 id="the-tools">The tools </h2><table> <thead> <tr> <th>Tool</th> <th>Provider</th> <th>Link</th> </tr> </thead> <tbody> <tr> <td>ChatGPT</td> <td>OpenAI</td> <td><a class="link" href="https://chat.openai.com" target="_blank" rel="noopener" >chat.openai.com</a></td> </tr> <tr> <td>Claude</td> <td>Anthropic</td> <td><a class="link" href="https://claude.ai" target="_blank" rel="noopener" >claude.ai</a></td> </tr> <tr> <td>Gemini</td> <td>Google</td> <td><a class="link" href="https://gemini.google.com" target="_blank" rel="noopener" >gemini.google.com</a></td> </tr> <tr> <td>Perplexity</td> <td>Perplexity AI</td> <td><a class="link" href="https://perplexity.ai" target="_blank" rel="noopener" >perplexity.ai</a></td> </tr> <tr> <td>Pi</td> <td>Inflection AI</td> <td><a class="link" href="https://pi.ai" target="_blank" rel="noopener" >pi.ai</a></td> </tr> </tbody> </table> <hr> <h2 id="what-to-use-for-what">What to use for what </h2><p><strong>ChatGPT</strong> — solid all-rounder. Good starting point for anything. GPT-4o handles code, writing, and general reasoning well. The memory and custom GPTs features are useful if you&rsquo;re in the ecosystem.</p> <p><strong>Claude</strong> — noticeably better at longer documents and editing. Tends to produce cleaner prose and is less likely to pad answers. Good choice when writing quality matters or when working with large amounts of text. The extended thinking mode (on paid plans) is useful for multi-step reasoning problems.</p> <p><strong>Gemini</strong> — largest context window of the group (1M tokens on the 1.5/2.0 Pro models). Good for dropping in entire codebases or large document sets. Also, the natural choice if you&rsquo;re already working in Google Workspace (Docs, Drive integration).</p> <p><strong>Perplexity</strong> — search-augmented: it retrieves and cites sources rather than relying solely on training data. Use it when you need something current, want to find specific information with references, or want to verify claims. Not the right tool for writing or code generation.</p> <p><strong>Pi</strong> — conversational in style, more of a back-and-forth dialogue companion. Less useful for technical tasks; more useful for thinking out loud or explaining something to yourself.</p> <hr> <h2 id="free-vs-paid">Free vs paid </h2><p>All of them have a free tier that covers most casual use. The paid tiers unlock better models (GPT-4o vs GPT-3.5, Claude Sonnet/Opus vs Haiku, etc.) and higher usage limits. If you&rsquo;re doing any serious work the free tier runs out quickly.</p> <p>For coding specifically, see <a class="link" href="https://backend-engineering-strategy-tools.github.io/site/public-notes/dev-environment/ai-coding/" >AI Coding Assistants</a> — the browser interfaces are fine for quick questions but the CLI tools are more useful when actually working in a codebase.</p>