Skip to content

an open-source alternative to the dots, bots and muses of the world, run an office of claude code/codex like agents on your laptop, sandboxes or anywhere, uses your existing subscriptions

seo-engine: rebuild blog for the Codex vs Claude Code refresh

chaitanyagiric3ba47b8165a
diff --git a/docs/blog/codex-cli-vs-claude-code/index.html b/docs/blog/codex-cli-vs-claude-code/index.html
index ccb55462..1c445633 100644
--- a/docs/blog/codex-cli-vs-claude-code/index.html
+++ b/docs/blog/codex-cli-vs-claude-code/index.html
@@ -18,8 +18,8 @@
<meta charset="UTF-8" />
<meta name="viewport" content="width=device-width, initial-scale=1.0" />
-<title>Codex CLI vs Claude Code — Munder Difflin Blog</title>
-<meta name="description" content="Codex CLI and Claude Code compared firsthand: sandbox and approval flags, AGENTS.md vs CLAUDE.md, real pricing, and where each terminal agent wins." />
+<title>Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026) — Munder Difflin Blog</title>
+<meta name="description" content="Codex vs Claude Code, checked 29 Sep 2026: models, plans, usage limits, sandboxing, computer use, MCP and subagents in one dated table, then which to pick." />
<link rel="canonical" href="https://munderdiffl.in/blog/codex-cli-vs-claude-code/" />
<link rel="icon" type="image/png" href="https://munderdiffl.in/logo.png" />
@@ -29,8 +29,8 @@
<!-- Open Graph -->
<meta property="og:site_name" content="Munder Difflin Blog" />
<meta property="og:type" content="article" />
-<meta property="og:title" content="Codex CLI vs Claude Code — Munder Difflin Blog" />
-<meta property="og:description" content="Codex CLI and Claude Code compared firsthand: sandbox and approval flags, AGENTS.md vs CLAUDE.md, real pricing, and where each terminal agent wins." />
+<meta property="og:title" content="Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026) — Munder Difflin Blog" />
+<meta property="og:description" content="Codex vs Claude Code, checked 29 Sep 2026: models, plans, usage limits, sandboxing, computer use, MCP and subagents in one dated table, then which to pick." />
<meta property="og:url" content="https://munderdiffl.in/blog/codex-cli-vs-claude-code/" />
<meta property="og:image" content="https://munderdiffl.in/blog/assets/media/codex-cli-vs-claude-code/hero.png" />
<meta property="og:image:width" content="1600" />
@@ -48,8 +48,8 @@
<!-- Twitter -->
<meta name="twitter:card" content="summary_large_image" />
-<meta name="twitter:title" content="Codex CLI vs Claude Code — Munder Difflin Blog" />
-<meta name="twitter:description" content="Codex CLI and Claude Code compared firsthand: sandbox and approval flags, AGENTS.md vs CLAUDE.md, real pricing, and where each terminal agent wins." />
+<meta name="twitter:title" content="Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026) — Munder Difflin Blog" />
+<meta name="twitter:description" content="Codex vs Claude Code, checked 29 Sep 2026: models, plans, usage limits, sandboxing, computer use, MCP and subagents in one dated table, then which to pick." />
<meta name="twitter:image" content="https://munderdiffl.in/blog/assets/media/codex-cli-vs-claude-code/hero.png" />
<meta name="twitter:image:alt" content="Illustration: Codex CLI vs Claude Code" />
@@ -63,11 +63,11 @@
{
"@context": "https://schema.org",
"@type": "BlogPosting",
- "headline": "Codex CLI vs Claude Code",
- "description": "Codex CLI and Claude Code compared firsthand: sandbox and approval flags, AGENTS.md vs CLAUDE.md, real pricing, and where each terminal agent wins.",
+ "headline": "Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026)",
+ "description": "Codex vs Claude Code, checked 29 Sep 2026: models, plans, usage limits, sandboxing, computer use, MCP and subagents in one dated table, then which to pick.",
"image": ["https://munderdiffl.in/blog/assets/media/codex-cli-vs-claude-code/hero.png"],
"datePublished": "2026-09-10T00:00:00.000Z",
- "dateModified": "2026-09-10T00:00:00.000Z",
+ "dateModified": "2026-09-29T00:00:00.000Z",
"author": { "@type": "Person", "name": "Chaitanya Giri" },
"publisher": {
"@type": "Organization",
@@ -75,7 +75,7 @@
"logo": { "@type": "ImageObject", "url": "https://munderdiffl.in/logo.png" }
},
"mainEntityOfPage": { "@type": "WebPage", "@id": "https://munderdiffl.in/blog/codex-cli-vs-claude-code/" },
- "keywords": "codex cli vs claude code, claude code vs codex cli, codex cli vs claude code comparison, codex cli approval mode, codex cli permission mode, does codex cli use agents md, codex cli vs claude code reddit",
+ "keywords": "codex vs claude code, claude code vs codex, codex cli vs claude code, codex vs claude code usage limits, codex vs claude code computer use, is codex cli free, does claude code read agents md",
"articleSection": "Comparisons",
"inLanguage": "en-US",
"url": "https://munderdiffl.in/blog/codex-cli-vs-claude-code/"
@@ -88,7 +88,7 @@
"itemListElement": [
{ "@type": "ListItem", "position": 1, "name": "Home", "item": "https://munderdiffl.in/" },
{ "@type": "ListItem", "position": 2, "name": "Blog", "item": "https://munderdiffl.in/blog/" },
- { "@type": "ListItem", "position": 3, "name": "Codex CLI vs Claude Code", "item": "https://munderdiffl.in/blog/codex-cli-vs-claude-code/" }
+ { "@type": "ListItem", "position": 3, "name": "Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026)", "item": "https://munderdiffl.in/blog/codex-cli-vs-claude-code/" }
]
}
</script>
@@ -98,11 +98,12 @@
"@context": "https://schema.org",
"@type": "FAQPage",
"mainEntity": [
- { "@type": "Question", "name": "Is Codex CLI free to use?", "acceptedAnswer": { "@type": "Answer", "text": "There is a real free tier, but it is capped. OpenAI's own Codex pricing page lists a Free plan at $0 a month for quick coding tasks, a Go plan at $8, and a Plus plan at $20 that adds cloud-based runs and the GPT-5.6 model family, checked 10 Sep 2026. Claude Code has no equivalent free tier: it authenticates against a paid Claude subscription or an Anthropic API key." } },
- { "@type": "Question", "name": "Does Codex CLI read AGENTS.md the way Claude Code reads CLAUDE.md?", "acceptedAnswer": { "@type": "Answer", "text": "Yes, but they are separate files, and each CLI's docs describe only its own. Codex CLI's docs list /init as the command that creates an AGENTS.md with instructions for Codex, checked 10 Sep 2026. Claude Code instead reads a project's CLAUDE.md at the start of every session, per Anthropic's own docs checked the same day, and neither doc promises to read the other's file." } },
- { "@type": "Question", "name": "What is the real difference between Codex CLI's sandbox and Claude Code's permission mode?", "acceptedAnswer": { "@type": "Answer", "text": "Codex CLI splits the decision in two: --sandbox (read-only, workspace-write, or danger-full-access) decides what it can touch, and --ask-for-approval (on-request or never) decides when it has to stop and ask, confirmed against codex --help on 10 Sep 2026. Claude Code folds both into one --permission-mode flag with six named modes, confirmed against claude --help the same day. Neither approach is wrong, Codex's is more composable and Claude Code's is a single dial to reason about." } },
- { "@type": "Question", "name": "Can I run Codex CLI and Claude Code together?", "acceptedAnswer": { "@type": "Answer", "text": "Yes. Munder Difflin wraps both as real terminal processes and runs Codex and Claude Code agents side by side on the same office floor, so each can take its own task at the same time. Munder Difflin 0.5.0 (2026-09-08) added GPT-6 Astra to Codex's model picker, and that slug needs Codex 0.153.1 or newer to be accepted when the agent spawns." } },
- { "@type": "Question", "name": "Codex CLI vs Claude Code, which one is actually better?", "acceptedAnswer": { "@type": "Answer", "text": "Neither wins outright. Codex CLI is the better pick when you want an open source binary you can audit, a free way in, and separate dials for sandboxing and approval. Claude Code is the better pick when you already pay for Claude and want a single permission mode, CLAUDE.md, and hooks that run around every tool call." } }
+ { "@type": "Question", "name": "Is Codex better than Claude Code?", "acceptedAnswer": { "@type": "Answer", "text": "Neither wins outright as of 29 Sep 2026. Codex is the better fit if you want to try it free in the desktop app, limits published per model, and computer use that can run in the background on a Mac. Claude Code is the better fit if you want Opus 5.5 as the default model and computer use from inside the CLI on a Pro or Max plan." } },
+ { "@type": "Question", "name": "Is Codex CLI free?", "acceptedAnswer": { "@type": "Answer", "text": "No. The CLI needs ChatGPT Plus ($20 a month) or higher, or an API key billed at API rates, per OpenAI's Codex pricing page checked 29 Sep 2026. The Free ($0) and Go ($8 a month) plans get GPT-6 Luna in the desktop app only, subject to rollout. Claude Code is not on Claude's Free plan either; it starts at Pro." } },
+ { "@type": "Question", "name": "Which has higher usage limits, Codex or Claude Code?", "acceptedAnswer": { "@type": "Answer", "text": "You can only compare them on paper for Codex, because OpenAI publishes estimated messages per five hours for each model and Anthropic does not. Both products meter a rolling five hour window plus a weekly limit, and both sell credits once you run out. The multipliers line up: 5x and 20x tiers on each side." } },
+ { "@type": "Question", "name": "Does Claude Code read AGENTS.md?", "acceptedAnswer": { "@type": "Answer", "text": "Yes, with a catch. Claude Code's memory docs, checked 29 Sep 2026, say it reads `AGENTS.md` by default only when there is no `CLAUDE.md`, `.claude/CLAUDE.md` or `CLAUDE.local.md` in the working directory or above it. If one exists, it reads the `CLAUDE.md` files only, unless that file imports `AGENTS.md` or you change the Project instructions setting. It needs Claude Code v2.1.277 or later." } },
+ { "@type": "Question", "name": "Does Codex CLI have computer use?", "acceptedAnswer": { "@type": "Answer", "text": "Not in the CLI. OpenAI's docs put Computer Use in the ChatGPT desktop app, with Codex, on macOS and Windows in supported regions, checked 29 Sep 2026. Claude Code has it inside the CLI, but only on macOS, as a research preview on Pro and Max plans." } },
+ { "@type": "Question", "name": "Can I run Codex and Claude Code together?", "acceptedAnswer": { "@type": "Answer", "text": "Yes. Munder Difflin, free and open source, runs Codex and Claude Code agents side by side on one office floor, each in its own terminal, and you choose the CLI per agent." } }
]
}
@@ -171,15 +172,15 @@
<span>/</span>
<a href="/blog/topics/comparisons/">Comparisons</a>
</nav>
- <h1>Codex CLI vs Claude Code</h1>
- <p class="dek">Codex CLI and Claude Code compared firsthand: sandbox and approval flags, AGENTS.md vs CLAUDE.md, real pricing, and where each terminal agent wins.</p>
+ <h1>Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026)</h1>
+ <p class="dek">Codex vs Claude Code, checked 29 Sep 2026: models, plans, usage limits, sandboxing, computer use, MCP and subagents in one dated table, then which to pick.</p>
<div class="byline">
<span class="av" aria-hidden="true">MD</span>
<strong>Chaitanya Giri</strong>
<span class="dot" aria-hidden="true">·</span>
<time datetime="2026-09-10">Sep 10, 2026</time>
<span class="dot" aria-hidden="true">·</span>
- <span>5 min read</span>
+ <span>7 min read</span>
<span class="tag kind">Comparisons</span>
</div>
</header>
@@ -188,32 +189,162 @@
<div class="post-wrap wrap">
<div class="prose">
- <p>Pick Codex CLI if you want an open source agent with a free plan and separate dials for sandboxing and approval. Pick Claude Code if you already pay for Claude and want one permission mode, <code>CLAUDE.md</code> and hooks around every tool call. Both run real commands in your repo; the difference is how each hands you control.</p>
-<p>Both are real terminal agents from frontier labs, and the differences that matter show up in how each one is built and how it hands you control, not in vague claims about which model is smarter. It’s the same lifecycle question we asked when comparing <a href="/blog/github-copilot-cli-vs-claude-code-for-hive-work/">GitHub Copilot CLI against Claude Code as hive workers</a>: which engine stays resident, and which one behaves the same way the second time you hand it control. If you’re weighing whole categories of <a href="/blog/best-claude-code-multi-agent-tools/">multi-agent tooling built around Claude Code</a> rather than just these two engines, start there instead. This post stays narrow: Codex CLI against Claude Code, checked on 10 Sep 2026.</p>
-<h2 id="whats-the-real-difference-between-codex-cli-and-claude-code" tabindex="-1">What’s the real difference between Codex CLI and Claude Code? <a class="anchor" href="#whats-the-real-difference-between-codex-cli-and-claude-code" aria-hidden="true">#</a></h2>
-<p>Codex CLI is an open-source, Rust-based binary from OpenAI that wraps each shell command in an OS-level sandbox and asks for approval on a separate policy. Claude Code is Anthropic’s CLI, and it folds sandboxing and approval into a single permission-mode setting instead of two.</p>
-<p>We checked this ourselves on 10 Sep 2026, with both CLIs already installed (<code>command -v claude</code> and <code>command -v codex</code> both resolved). <code>codex --help</code> lists <code>-s, --sandbox &lt;SANDBOX_MODE&gt;</code> with three possible values (<code>read-only</code>, <code>workspace-write</code>, <code>danger-full-access</code>), and a separate <code>-a, --ask-for-approval &lt;APPROVAL_POLICY&gt;</code> (<code>on-request</code> or <code>never</code>). <code>claude --help</code> lists one <code>--permission-mode &lt;mode&gt;</code> flag instead, with six named modes: <code>acceptEdits</code>, <code>auto</code>, <code>bypassPermissions</code>, <code>manual</code>, <code>dontAsk</code>, <code>plan</code>. Same underlying job, two different shapes: Codex separates “what it’s allowed to touch” from “when it must ask,” Claude Code bundles both into one choice.</p>
-<p>OpenAI’s own docs describe <code>workspace-write</code> as the mode where Codex “can read files, make edits, and run commands in the workspace,” and <code>on-request</code>, the policy the CLI runs by default, as the one where Codex “requires approval to edit outside the workspace or to access network” and otherwise carries on (<a href="https://developers.openai.com/codex/agent-approvals-security">developers.openai.com/codex/agent-approvals-security</a>, checked 10 Sep 2026). Both projects also named their full escape hatch the same way: Codex’s is <code>--dangerously-bypass-approvals-and-sandbox</code>, Claude Code’s is <code>--dangerously-skip-permissions</code>. Somebody in both API design rooms clearly wanted you to feel that flag before you typed it, which is about as close as a CLI flag gets to a performance review.</p>
-<h2 id="does-codex-cli-read-agentsmd-the-way-claude-code-reads-claudemd" tabindex="-1">Does Codex CLI read <code>AGENTS.md</code> the way Claude Code reads <code>CLAUDE.md</code>? <a class="anchor" href="#does-codex-cli-read-agentsmd-the-way-claude-code-reads-claudemd" aria-hidden="true">#</a></h2>
-<p>Yes, but the two files are separate, and each CLI’s docs describe only its own. Codex CLI’s docs list <code>/init</code> as the command to “create an <code>AGENTS.md</code> file with instructions for Codex” (<a href="https://developers.openai.com/codex/cli/">developers.openai.com/codex/cli</a>, checked 10 Sep 2026). Claude Code instead reads a project’s <code>CLAUDE.md</code> at the start of every session, described by Anthropic as “a markdown file you add to your project root that Claude Code reads at the start of every session” (<a href="https://code.claude.com/docs/en/overview">code.claude.com/docs/en/overview</a>, checked 10 Sep 2026).</p>
-<p>The two overlap heavily in practice: coding standards, architecture notes, a review checklist. What differs is delivery. Neither CLI’s docs promise to read the other’s file, so a repo that runs both engines usually keeps both <code>AGENTS.md</code> and <code>CLAUDE.md</code>, or keeps one and symlinks the other to it.</p>
+ <p>Codex and Claude Code are close enough in September 2026 that your plan should decide: pick Codex for a free desktop tier, limits published per model and background computer use on a Mac; pick Claude Code for Opus 5.5 by default and computer use inside the terminal.</p>
+<p>You can also skip the choice: switch between the two by hand, or run both in <a href="https://harnessmd.com/download">Munder Difflin</a>, free and open source, which puts Codex and Claude Code agents side by side on one floor, each in its own terminal. Everything below was checked on 29 Sep 2026 against each vendor’s own pricing pages and docs.</p>
+<h2 id="codex-vs-claude-code-at-a-glance-checked-29-sep-2026" tabindex="-1">Codex vs Claude Code at a glance (checked 29 Sep 2026) <a class="anchor" href="#codex-vs-claude-code-at-a-glance-checked-29-sep-2026" aria-hidden="true">#</a></h2>
+<div class="table-scroll">
+<table>
+<thead>
+<tr>
+<th style="text-align:left"></th>
+<th style="text-align:left">Codex (OpenAI)</th>
+<th style="text-align:left">Claude Code (Anthropic)</th>
+</tr>
+</thead>
+<tbody>
+<tr>
+<td style="text-align:left">Where it runs</td>
+<td style="text-align:left">CLI, IDE extension, ChatGPT desktop app, web, iOS</td>
+<td style="text-align:left">CLI, IDE extensions, desktop app, web</td>
+</tr>
+<tr>
+<td style="text-align:left">Source</td>
+<td style="text-align:left">CLI is Apache-2.0 on GitHub</td>
+<td style="text-align:left">Closed, Anthropic commercial terms</td>
+</tr>
+<tr>
+<td style="text-align:left">Models</td>
+<td style="text-align:left">GPT-6 Astra, GPT-6 Sol, GPT-6 Luna, GPT-5.6 family. The pricing page still lists GPT-5.5 (retires 14 Oct 2026), GPT-5.4 and GPT-5.4 mini</td>
+<td style="text-align:left">Opus 5.5 is the default on Pro, Max, Team and Enterprise, plus Sonnet and Haiku. Fable 5.1 can bill to usage credits, depending on plan and seat tier</td>
+</tr>
+<tr>
+<td style="text-align:left">Cheapest plan with the CLI</td>
+<td style="text-align:left">Plus, $20 a month (Free and Go get the desktop app only)</td>
+<td style="text-align:left">Pro, $20 a month</td>
+</tr>
+<tr>
+<td style="text-align:left">Heavy plans</td>
+<td style="text-align:left">Pro 5x from $100, Pro 20x at $200 (new sign ups paused, reported 11 Sep 2026)</td>
+<td style="text-align:left">Max 5x at $100, Max 20x at $200</td>
+</tr>
+<tr>
+<td style="text-align:left">Limits</td>
+<td style="text-align:left">Published ranges per model per 5 hours, weekly limits may apply, credits after</td>
+<td style="text-align:left">5 hour session limit plus a weekly limit, no published message counts, credits after</td>
+</tr>
+<tr>
+<td style="text-align:left">Sandbox</td>
+<td style="text-align:left">On by default: Seatbelt on macOS, bubblewrap on Linux and WSL2, native Windows sandbox</td>
+<td style="text-align:left">Built in Bash sandbox (Seatbelt, bubblewrap), set up with <code>/sandbox</code>. No native Windows</td>
+</tr>
+<tr>
+<td style="text-align:left">Approval controls</td>
+<td style="text-align:left"><code>--sandbox</code> and <code>--ask-for-approval</code>, two flags</td>
+<td style="text-align:left">One <code>--permission-mode</code> with six modes</td>
+</tr>
+<tr>
+<td style="text-align:left">Computer use</td>
+<td style="text-align:left">ChatGPT desktop app on macOS and Windows, background on macOS, not in the CLI</td>
+<td style="text-align:left">In the CLI on macOS, research preview, Pro and Max only. Desktop app: macOS and Windows</td>
+</tr>
+<tr>
+<td style="text-align:left">MCP</td>
+<td style="text-align:left">Yes, STDIO and streamable HTTP</td>
+<td style="text-align:left">Yes, local stdio and remote HTTP</td>
+</tr>
+<tr>
+<td style="text-align:left">Subagents</td>
+<td style="text-align:left">On by default in current releases, plus custom agents</td>
+<td style="text-align:left">Built in and custom, each in its own context window</td>
+</tr>
+<tr>
+<td style="text-align:left">Project instructions</td>
+<td style="text-align:left"><code>AGENTS.md</code></td>
+<td style="text-align:left"><code>CLAUDE.md</code>. Reads <code>AGENTS.md</code> only when no <code>CLAUDE.md</code> exists, by default</td>
+</tr>
+</tbody>
+</table>
+</div>
+<p>Sources: OpenAI’s <a href="https://learn.chatgpt.com/docs/pricing">Codex pricing page</a>, Anthropic’s <a href="https://support.claude.com/en/articles/11049741-what-is-the-max-plan">Max plan article</a> and Claude Code’s model configuration docs, all checked 29 Sep 2026. Sandbox flags come from <code>codex --help</code> and <code>claude --help</code>, run the same day.</p>
+<h2 id="which-one-has-higher-usage-limits-codex-or-claude-code" tabindex="-1">Which one has higher usage limits, Codex or Claude Code? <a class="anchor" href="#which-one-has-higher-usage-limits-codex-or-claude-code" aria-hidden="true">#</a></h2>
+<p>Only Codex lets you compare on paper, because OpenAI publishes estimated local messages per five hours for each model and Anthropic publishes none. OpenAI’s pricing page, checked 29 Sep 2026, lists these ranges:</p>
+<div class="table-scroll">
+<table>
+<thead>
+<tr>
+<th style="text-align:left">Model</th>
+<th style="text-align:left">Plus</th>
+<th style="text-align:left">Pro 5x</th>
+<th style="text-align:left">Pro 20x</th>
+</tr>
+</thead>
+<tbody>
+<tr>
+<td style="text-align:left">GPT-6 Astra</td>
+<td style="text-align:left">5 to 45</td>
+<td style="text-align:left">25 to 225</td>
+<td style="text-align:left">100 to 900</td>
+</tr>
+<tr>
+<td style="text-align:left">GPT-6 Sol</td>
+<td style="text-align:left">15 to 150</td>
+<td style="text-align:left">70 to 700</td>
+<td style="text-align:left">300 to 3,000</td>
+</tr>
+<tr>
+<td style="text-align:left">GPT-6 Luna</td>
+<td style="text-align:left">350 to 3,000</td>
+<td style="text-align:left">1,750 to 14,000</td>
+<td style="text-align:left">7,000 to 56,000</td>
+</tr>
+</tbody>
+</table>
+</div>
+<p>The same page says weekly limits may also apply, and that Plus and Pro users can buy credits when they run out. The spread inside each range comes from task size, context and reasoning, so a long Astra session lands near the bottom end.</p>
+<p>Anthropic describes its limits as multipliers instead. Max 5x gives five times Pro’s per session usage and Max 20x gives twenty times, both with a weekly limit across all models that resets at a fixed time for your account, per its Max plan article checked 29 Sep 2026. Anthropic’s Opus 5.5 announcement on 22 Sep 2026 also raised five hour limits on Pro, Max, Team and seat based Enterprise. Claude Code and the Claude apps share one pool, so a long chat eats into your coding time. Our <a href="/blog/claude-code-max-plan-tips/">Claude Code Max plan tips</a> cover how to stretch it, and the <a href="/blog/codex-max-plan-tips/">Codex Max plan tips</a> do the same for OpenAI’s side.</p>
+<h3 id="can-new-subscribers-still-get-chatgpt-pro-20x" tabindex="-1">Can new subscribers still get ChatGPT Pro 20x? <a class="anchor" href="#can-new-subscribers-still-get-chatgpt-pro-20x" aria-hidden="true">#</a></h3>
+<p>Probably not right now. <a href="https://fortune.com/2026/09/11/openai-astra-chatgpt-pro-pause/">Fortune reported on 11 Sep 2026</a> that OpenAI had paused new sign ups and upgrades to the $200 ChatGPT Pro tier because of demand for GPT-6 Astra, with existing Pro accounts unaffected. We could not load OpenAI’s own help page to confirm it, so check before you plan around it. If you burn through Pro 5x, your options are credits, an API key, or a second tool. Anthropic’s Max plan article carried no pause of that kind on 29 Sep 2026.</p>
+<h2 id="does-codex-or-claude-code-have-better-computer-use" tabindex="-1">Does Codex or Claude Code have better computer use? <a class="anchor" href="#does-codex-or-claude-code-have-better-computer-use" aria-hidden="true">#</a></h2>
+<p>Codex does more on a Mac, and Claude Code is the only one of the two that does it from the terminal. OpenAI’s <a href="https://learn.chatgpt.com/docs/computer-use">Computer Use docs</a>, checked 29 Sep 2026, put the feature in the ChatGPT desktop app with Codex, on macOS and Windows, in supported regions. On macOS it can run a scoped task in the background while you keep working elsewhere. On Windows it takes over the active desktop.</p>
+<p>Claude Code’s <a href="https://code.claude.com/docs/en/computer-use">computer use docs</a>, checked the same day, describe a built in <code>computer-use</code> MCP server you enable from <code>/mcp</code>. It is a macOS research preview for Pro and Max plans only, not Team or Enterprise, and it will not run under <code>-p</code>. Only one session can hold the screen at a time, and other apps are hidden while Claude works. If you want an agent clicking through your app while you type in another window, Codex’s desktop app is the better fit. If you want the agent that just wrote the Swift to launch it and click every button in the same terminal session, that is Claude Code.</p>
+<h2 id="how-do-sandboxing-and-approvals-differ" tabindex="-1">How do sandboxing and approvals differ? <a class="anchor" href="#how-do-sandboxing-and-approvals-differ" aria-hidden="true">#</a></h2>
+<p>Codex splits the decision into two flags and Claude Code folds it into one. We ran both help screens on 29 Sep 2026 with <code>codex-cli 0.153.4</code> and Claude Code <code>2.1.284</code> installed:</p>
+<pre class="language-text"><code class="language-text">codex --help
+ -s, --sandbox <SANDBOX_MODE>
+ [possible values: read-only, workspace-write, danger-full-access]
+ -a, --ask-for-approval <APPROVAL_POLICY>
+ - on-request: The model decides when to ask the user for approval
+ - never: Never ask for user approval
+
+claude --help
+ --permission-mode <mode> (choices: "acceptEdits", "auto",
+ "bypassPermissions", "manual",
+ "dontAsk", "plan")</code></pre>
+<p>Codex applies its sandbox in the default permissions mode, so every command it spawns, <code>git</code> and test runners included, inherits the boundary. Claude Code’s Bash sandbox uses the same OS primitives (Seatbelt on macOS, bubblewrap on Linux and WSL2) but you set it up with <code>/sandbox</code>. Both kept a loud escape hatch, <code>--dangerously-bypass-approvals-and-sandbox</code> and <code>--dangerously-skip-permissions</code>. Somebody in both design rooms wanted you to feel that flag before you typed it.</p>
<figure class="fig fig-inline"><img src="/blog/assets/media/codex-cli-vs-claude-code/note-1.png" alt="Hand-drawn sketch from “Codex CLI vs Claude Code”" loading="lazy" decoding="async" /></figure>
-<h2 id="is-codex-cli-free-to-use" tabindex="-1">Is Codex CLI free to use? <a class="anchor" href="#is-codex-cli-free-to-use" aria-hidden="true">#</a></h2>
-<p>There’s a real free tier, but it’s capped. OpenAI’s Codex pricing page lists a Free plan at $0 a month for exploring Codex “on quick coding tasks,” a Go plan at $8 a month, and a Plus plan at $20 a month that adds cloud-based runs and the GPT-5.6 model family (<a href="https://developers.openai.com/codex/pricing">developers.openai.com/codex/pricing</a>, checked 10 Sep 2026). Claude Code has no equivalent free tier of its own: it authenticates against a paid Claude subscription or an Anthropic API key from the first session. If you already pay for ChatGPT Pro, see <a href="/blog/codex-max-plan-tips/">how to make the most of your Codex Max plan</a>.</p>
-<h2 id="where-does-codex-cli-actually-win" tabindex="-1">Where does Codex CLI actually win? <a class="anchor" href="#where-does-codex-cli-actually-win" aria-hidden="true">#</a></h2>
-<p>Two places: openness and cost of entry. The CLI itself is Apache-2.0 licensed and public on GitHub, so you can read exactly what the sandbox and approval code does instead of trusting a closed binary (<a href="https://github.com/openai/codex">github.com/openai/codex</a>, checked 10 Sep 2026), and the free plan above gets a curious developer to a real sandboxed agent at no cost. If you want a run that never pauses to ask, <code>--ask-for-approval never</code> is also a cleaner single flag than composing Claude Code’s mode and its skip flag together. None of that makes Codex CLI the better tool in general. It makes it the better tool when auditability and a no-cost start matter most.</p>
+<h2 id="does-claude-code-read-agentsmd" tabindex="-1">Does Claude Code read <code>AGENTS.md</code>? <a class="anchor" href="#does-claude-code-read-agentsmd" aria-hidden="true">#</a></h2>
+<p>Yes, but only when there is no <code>CLAUDE.md</code>, by default. Claude Code’s memory docs, checked 29 Sep 2026, say it reads <code>AGENTS.md</code> when there is no <code>CLAUDE.md</code>, <code>.claude/CLAUDE.md</code> or <code>CLAUDE.local.md</code> in your working directory or above it. If one of those exists, it reads your <code>CLAUDE.md</code> files only. To share one file, import <code>AGENTS.md</code> from <code>CLAUDE.md</code>, or change the Project instructions setting to read both. Direct reading needs Claude Code v2.1.277 or later. Codex uses <code>AGENTS.md</code> natively and generates one with <code>/init</code>, so a repo with only an <code>AGENTS.md</code> now briefs both agents.</p>
+<h2 id="is-codex-cli-free" tabindex="-1">Is Codex CLI free? <a class="anchor" href="#is-codex-cli-free" aria-hidden="true">#</a></h2>
+<p>No: the free tier covers the desktop app, not the CLI. OpenAI’s pricing page, checked 29 Sep 2026, lists Free at $0 and Go at $8 a month, both with GPT-6 Luna in the desktop app subject to rollout, and Plus at $20 a month for Codex on the web, in the CLI, in the IDE extension and on iOS. You can also run the CLI on an API key and pay API rates. Claude Code is not on Claude’s Free plan; it starts at Pro.</p>
<figure class="fig fig-inline"><img src="/blog/assets/media/codex-cli-vs-claude-code/note-2.png" alt="Hand-drawn sketch from “Codex CLI vs Claude Code”" loading="lazy" decoding="async" /></figure>
-<h2 id="where-does-claude-code-actually-win" tabindex="-1">Where does Claude Code actually win? <a class="anchor" href="#where-does-claude-code-actually-win" aria-hidden="true">#</a></h2>
-<p>Two places: one engine across the tools you already use, and hooks. The terminal, VS Code, JetBrains, the desktop
-app and the web all run the same Claude Code engine, so a repo’s <code>CLAUDE.md</code>, settings and MCP servers behave the same
-wherever you open it (<a href="https://code.claude.com/docs/en/overview">code.claude.com/docs/en/overview</a>, checked 10 Sep
-2026). Hooks run shell commands before or after Claude Code’s own actions, like formatting after every file edit or
-running lint before a commit, so a rule you write once holds for every session instead of living in a README nobody
-rereads. If you already pay for Claude, that is a lot of control for no extra setup.</p>
-<p>Weighing Google’s agent as well? Read <a href="/blog/claude-code-vs-antigravity/">Claude Code vs Antigravity</a>.</p>
-<h2 id="can-you-run-codex-cli-and-claude-code-together" tabindex="-1">Can you run Codex CLI and Claude Code together? <a class="anchor" href="#can-you-run-codex-cli-and-claude-code-together" aria-hidden="true">#</a></h2>
-<p>Yes, and running both inside one harness is often more useful than picking a single winner. Munder Difflin wraps CLIs like <code>claude</code> and <code>codex</code> as real processes in their own terminals and runs them side by side on the same office floor, so a Codex agent and a Claude Code agent can each work on their own task at the same time. Munder Difflin 0.5.0 (2026-09-08) added GPT-6 Astra to Codex’s model picker, and that slug needs Codex 0.153.1 or newer to be accepted when the agent spawns; the Codex CLI checked for this piece, <code>codex-cli 0.153.4</code>, clears that bar.</p>
-<p>For the wider field beyond these two, the <a href="/blog/best-ai-coding-agents/">2026 field guide to AI coding agents</a> covers where in-editor and autonomous agents fit next to CLI ones. For this specific pair, the honest answer holds: reach for Claude Code when you already pay for Claude and want hooks and one permission mode, reach for Codex CLI when you want separate sandbox and approval dials and a free way to start, and don’t feel obligated to pick just one.</p>
+<h2 id="pick-codex-if-pick-claude-code-if" tabindex="-1">Pick Codex if…, pick Claude Code if… <a class="anchor" href="#pick-codex-if-pick-claude-code-if" aria-hidden="true">#</a></h2>
+<p><strong>Pick Codex if:</strong></p>
+<ul>
+<li>You want to try Codex free in the desktop app before you pay.</li>
+<li>You like knowing roughly how many messages each model buys you per five hours.</li>
+<li>You want computer use that runs in the background on a Mac.</li>
+<li>You want to read the CLI’s source (Apache-2.0).</li>
+</ul>
+<p><strong>Pick Claude Code if:</strong></p>
+<ul>
+<li>You want Opus 5.5 as your default model without touching settings.</li>
+<li>You want the agent to drive native apps and simulators from the same terminal session.</li>
+<li>You already pay for Claude and want chat and coding on one plan.</li>
+<li>You want one permission dial instead of two, and a sandbox you switch on per project.</li>
+</ul>
+<p>If you are still choosing between whole categories of tools, not just these two engines, the <a href="/blog/best-claude-code-multi-agent-tools/">multi-agent tools roundup</a> is the wider map, and <a href="/blog/claude-code-vs-antigravity/">Claude Code vs Antigravity</a> covers Google’s agent.</p>
+<h2 id="can-you-run-codex-and-claude-code-together" tabindex="-1">Can you run Codex and Claude Code together? <a class="anchor" href="#can-you-run-codex-and-claude-code-together" aria-hidden="true">#</a></h2>
+<p>Yes, and for a lot of teams that is the honest answer. Munder Difflin (as of 0.5.3) lets you pick the CLI per agent, Claude Code or Codex among others, and runs each one as a real terminal process on the same floor. A Codex agent can take the UI bug while a Claude Code agent takes the refactor. When one plan hits its weekly limit, the other keeps working.</p>
@@ -221,28 +352,33 @@ rereads. If you already pay for Claude, that is a lot of control for no extra se
<h2 id="faq">FAQ</h2>
<div class="qa">
- <p class="q">Is Codex CLI free to use?</p>
- <p class="a">There is a real free tier, but it is capped. OpenAI&#39;s own Codex pricing page lists a Free plan at $0 a month for quick coding tasks, a Go plan at $8, and a Plus plan at $20 that adds cloud-based runs and the GPT-5.6 model family, checked 10 Sep 2026. Claude Code has no equivalent free tier: it authenticates against a paid Claude subscription or an Anthropic API key.</p>
+ <p class="q">Is Codex better than Claude Code?</p>
+ <p class="a">Neither wins outright as of 29 Sep 2026. Codex is the better fit if you want to try it free in the desktop app, limits published per model, and computer use that can run in the background on a Mac. Claude Code is the better fit if you want Opus 5.5 as the default model and computer use from inside the CLI on a Pro or Max plan.</p>
+ </div>
+
+ <div class="qa">
+ <p class="q">Is Codex CLI free?</p>
+ <p class="a">No. The CLI needs ChatGPT Plus ($20 a month) or higher, or an API key billed at API rates, per OpenAI&#39;s Codex pricing page checked 29 Sep 2026. The Free ($0) and Go ($8 a month) plans get GPT-6 Luna in the desktop app only, subject to rollout. Claude Code is not on Claude&#39;s Free plan either; it starts at Pro.</p>
</div>
<div class="qa">
- <p class="q">Does Codex CLI read AGENTS.md the way Claude Code reads CLAUDE.md?</p>
- <p class="a">Yes, but they are separate files, and each CLI&#39;s docs describe only its own. Codex CLI&#39;s docs list /init as the command that creates an AGENTS.md with instructions for Codex, checked 10 Sep 2026. Claude Code instead reads a project&#39;s CLAUDE.md at the start of every session, per Anthropic&#39;s own docs checked the same day, and neither doc promises to read the other&#39;s file.</p>
+ <p class="q">Which has higher usage limits, Codex or Claude Code?</p>
+ <p class="a">You can only compare them on paper for Codex, because OpenAI publishes estimated messages per five hours for each model and Anthropic does not. Both products meter a rolling five hour window plus a weekly limit, and both sell credits once you run out. The multipliers line up: 5x and 20x tiers on each side.</p>
</div>
<div class="qa">
- <p class="q">What is the real difference between Codex CLI&#39;s sandbox and Claude Code&#39;s permission mode?</p>
- <p class="a">Codex CLI splits the decision in two: --sandbox (read-only, workspace-write, or danger-full-access) decides what it can touch, and --ask-for-approval (on-request or never) decides when it has to stop and ask, confirmed against codex --help on 10 Sep 2026. Claude Code folds both into one --permission-mode flag with six named modes, confirmed against claude --help the same day. Neither approach is wrong, Codex&#39;s is more composable and Claude Code&#39;s is a single dial to reason about.</p>
+ <p class="q">Does Claude Code read AGENTS.md?</p>
+ <p class="a">Yes, with a catch. Claude Code&#39;s memory docs, checked 29 Sep 2026, say it reads `AGENTS.md` by default only when there is no `CLAUDE.md`, `.claude/CLAUDE.md` or `CLAUDE.local.md` in the working directory or above it. If one exists, it reads the `CLAUDE.md` files only, unless that file imports `AGENTS.md` or you change the Project instructions setting. It needs Claude Code v2.1.277 or later.</p>
</div>
<div class="qa">
- <p class="q">Can I run Codex CLI and Claude Code together?</p>
- <p class="a">Yes. Munder Difflin wraps both as real terminal processes and runs Codex and Claude Code agents side by side on the same office floor, so each can take its own task at the same time. Munder Difflin 0.5.0 (2026-09-08) added GPT-6 Astra to Codex&#39;s model picker, and that slug needs Codex 0.153.1 or newer to be accepted when the agent spawns.</p>
+ <p class="q">Does Codex CLI have computer use?</p>
+ <p class="a">Not in the CLI. OpenAI&#39;s docs put Computer Use in the ChatGPT desktop app, with Codex, on macOS and Windows in supported regions, checked 29 Sep 2026. Claude Code has it inside the CLI, but only on macOS, as a research preview on Pro and Max plans.</p>
</div>
<div class="qa">
- <p class="q">Codex CLI vs Claude Code, which one is actually better?</p>
- <p class="a">Neither wins outright. Codex CLI is the better pick when you want an open source binary you can audit, a free way in, and separate dials for sandboxing and approval. Claude Code is the better pick when you already pay for Claude and want a single permission mode, CLAUDE.md, and hooks that run around every tool call.</p>
+ <p class="q">Can I run Codex and Claude Code together?</p>
+ <p class="a">Yes. Munder Difflin, free and open source, runs Codex and Claude Code agents side by side on one office floor, each in its own terminal, and you choose the CLI per agent.</p>
</div>
</section>
@@ -261,7 +397,7 @@ rereads. If you already pay for Claude, that is a lot of control for no extra se
<aside class="toc-rail" aria-label="Table of contents">
<div class="tt">On this page</div>
<ol>
- <li class="lvl-2"><a href="#whats-the-real-difference-between-codex-cli-and-claude-code">What’s the real difference between Codex CLI and Claude Code?</a></li><li class="lvl-2"><a href="#does-codex-cli-read-agentsmd-the-way-claude-code-reads-claudemd">Does Codex CLI read AGENTS.md the way Claude Code reads CLAUDE.md?</a></li><li class="lvl-2"><a href="#is-codex-cli-free-to-use">Is Codex CLI free to use?</a></li><li class="lvl-2"><a href="#where-does-codex-cli-actually-win">Where does Codex CLI actually win?</a></li><li class="lvl-2"><a href="#where-does-claude-code-actually-win">Where does Claude Code actually win?</a></li><li class="lvl-2"><a href="#can-you-run-codex-cli-and-claude-code-together">Can you run Codex CLI and Claude Code together?</a></li>
+ <li class="lvl-2"><a href="#codex-vs-claude-code-at-a-glance-checked-29-sep-2026">Codex vs Claude Code at a glance (checked 29 Sep 2026)</a></li><li class="lvl-2"><a href="#which-one-has-higher-usage-limits-codex-or-claude-code">Which one has higher usage limits, Codex or Claude Code?</a></li><li class="lvl-3"><a href="#can-new-subscribers-still-get-chatgpt-pro-20x">Can new subscribers still get ChatGPT Pro 20x?</a></li><li class="lvl-2"><a href="#does-codex-or-claude-code-have-better-computer-use">Does Codex or Claude Code have better computer use?</a></li><li class="lvl-2"><a href="#how-do-sandboxing-and-approvals-differ">How do sandboxing and approvals differ?</a></li><li class="lvl-2"><a href="#does-claude-code-read-agentsmd">Does Claude Code read AGENTS.md?</a></li><li class="lvl-2"><a href="#is-codex-cli-free">Is Codex CLI free?</a></li><li class="lvl-2"><a href="#pick-codex-if-pick-claude-code-if">Pick Codex if…, pick Claude Code if…</a></li><li class="lvl-2"><a href="#can-you-run-codex-and-claude-code-together">Can you run Codex and Claude Code together?</a></li>
</ol>
</aside>
diff --git a/docs/blog/feed.xml b/docs/blog/feed.xml
index fc251006..2a8756f6 100644
--- a/docs/blog/feed.xml
+++ b/docs/blog/feed.xml
@@ -216,12 +216,12 @@
<summary>An MCP server gives an AI agent tools, data and prompts over one open protocol. What it is, the host, client and server roles, and where Claude Code plugs in.</summary>
</entry>
<entry>
- <title>Codex CLI vs Claude Code</title>
+ <title>Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026)</title>
<link href="https://munderdiffl.in/blog/codex-cli-vs-claude-code/" />
<id>https://munderdiffl.in/blog/codex-cli-vs-claude-code/</id>
<published>2026-09-10T00:00:00.000Z</published>
- <updated>2026-09-10T00:00:00.000Z</updated><category term="Comparisons" />
- <summary>Codex CLI and Claude Code compared firsthand: sandbox and approval flags, AGENTS.md vs CLAUDE.md, real pricing, and where each terminal agent wins.</summary>
+ <updated>2026-09-29T00:00:00.000Z</updated><category term="Comparisons" />
+ <summary>Codex vs Claude Code, checked 29 Sep 2026: models, plans, usage limits, sandboxing, computer use, MCP and subagents in one dated table, then which to pick.</summary>
</entry>
<entry>
<title>How Much Does Claude Code Cost?</title>
diff --git a/docs/blog/how-much-does-claude-code-cost/index.html b/docs/blog/how-much-does-claude-code-cost/index.html
index 7125b27f..af50a55c 100644
--- a/docs/blog/how-much-does-claude-code-cost/index.html
+++ b/docs/blog/how-much-does-claude-code-cost/index.html
@@ -270,7 +270,7 @@
<footer class="post-foot wrap"><div class="prevnext">
<a class="pn prev" href="/blog/how-to-add-an-mcp-server-to-claude-code/"><span class="k">← Older</span><span class="t">How to Add an MCP Server to Claude Code</span></a>
- <a class="pn next" href="/blog/codex-cli-vs-claude-code/"><span class="k">Newer →</span><span class="t">Codex CLI vs Claude Code</span></a>
+ <a class="pn next" href="/blog/codex-cli-vs-claude-code/"><span class="k">Newer →</span><span class="t">Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026)</span></a>
</div>
<section class="related">
<h2>Related in Concepts</h2>
diff --git a/docs/blog/index.html b/docs/blog/index.html
index 56b0bd05..e4205eb5 100644
--- a/docs/blog/index.html
+++ b/docs/blog/index.html
@@ -624,13 +624,13 @@
<div class="meta">
<time datetime="2026-09-10">Sep 10, 2026</time>
<span class="dot" aria-hidden="true"></span>
- <span>5 min</span>
+ <span>7 min</span>
</div>
- <h3><a href="/blog/codex-cli-vs-claude-code/">Codex CLI vs Claude Code</a></h3>
- <p class="dek">Codex CLI and Claude Code compared firsthand: sandbox and approval flags, AGENTS.md vs CLAUDE.md, real pricing, and where each terminal agent wins.</p>
+ <h3><a href="/blog/codex-cli-vs-claude-code/">Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026)</a></h3>
+ <p class="dek">Codex vs Claude Code, checked 29 Sep 2026: models, plans, usage limits, sandboxing, computer use, MCP and subagents in one dated table, then which to pick.</p>
<div class="foot">
<span class="tag kind">Comparisons</span>
- <a class="more" href="/blog/codex-cli-vs-claude-code/" aria-label="Read: Codex CLI vs Claude Code">Read →</a>
+ <a class="more" href="/blog/codex-cli-vs-claude-code/" aria-label="Read: Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026)">Read →</a>
</div>
</div>
</article>
diff --git a/docs/blog/tags/claude-code/index.html b/docs/blog/tags/claude-code/index.html
index fc76e0cd..1263a164 100644
--- a/docs/blog/tags/claude-code/index.html
+++ b/docs/blog/tags/claude-code/index.html
@@ -355,13 +355,13 @@
<div class="meta">
<time datetime="2026-09-10">Sep 10, 2026</time>
<span class="dot" aria-hidden="true"></span>
- <span>5 min</span>
+ <span>7 min</span>
</div>
- <h3><a href="/blog/codex-cli-vs-claude-code/">Codex CLI vs Claude Code</a></h3>
- <p class="dek">Codex CLI and Claude Code compared firsthand: sandbox and approval flags, AGENTS.md vs CLAUDE.md, real pricing, and where each terminal agent wins.</p>
+ <h3><a href="/blog/codex-cli-vs-claude-code/">Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026)</a></h3>
+ <p class="dek">Codex vs Claude Code, checked 29 Sep 2026: models, plans, usage limits, sandboxing, computer use, MCP and subagents in one dated table, then which to pick.</p>
<div class="foot">
<span class="tag kind">Comparisons</span>
- <a class="more" href="/blog/codex-cli-vs-claude-code/" aria-label="Read: Codex CLI vs Claude Code">Read →</a>
+ <a class="more" href="/blog/codex-cli-vs-claude-code/" aria-label="Read: Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026)">Read →</a>
</div>
</div>
</article>
diff --git a/docs/blog/tags/codex/index.html b/docs/blog/tags/codex/index.html
index b84290f3..a0db90a8 100644
--- a/docs/blog/tags/codex/index.html
+++ b/docs/blog/tags/codex/index.html
@@ -147,13 +147,13 @@
<div class="meta">
<time datetime="2026-09-10">Sep 10, 2026</time>
<span class="dot" aria-hidden="true"></span>
- <span>5 min</span>
+ <span>7 min</span>
</div>
- <h3><a href="/blog/codex-cli-vs-claude-code/">Codex CLI vs Claude Code</a></h3>
- <p class="dek">Codex CLI and Claude Code compared firsthand: sandbox and approval flags, AGENTS.md vs CLAUDE.md, real pricing, and where each terminal agent wins.</p>
+ <h3><a href="/blog/codex-cli-vs-claude-code/">Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026)</a></h3>
+ <p class="dek">Codex vs Claude Code, checked 29 Sep 2026: models, plans, usage limits, sandboxing, computer use, MCP and subagents in one dated table, then which to pick.</p>
<div class="foot">
<span class="tag kind">Comparisons</span>
- <a class="more" href="/blog/codex-cli-vs-claude-code/" aria-label="Read: Codex CLI vs Claude Code">Read →</a>
+ <a class="more" href="/blog/codex-cli-vs-claude-code/" aria-label="Read: Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026)">Read →</a>
</div>
</div>
</article>
diff --git a/docs/blog/tags/comparisons/index.html b/docs/blog/tags/comparisons/index.html
index 3cf11c9a..67048a18 100644
--- a/docs/blog/tags/comparisons/index.html
+++ b/docs/blog/tags/comparisons/index.html
@@ -291,13 +291,13 @@
<div class="meta">
<time datetime="2026-09-10">Sep 10, 2026</time>
<span class="dot" aria-hidden="true"></span>
- <span>5 min</span>
+ <span>7 min</span>
</div>
- <h3><a href="/blog/codex-cli-vs-claude-code/">Codex CLI vs Claude Code</a></h3>
- <p class="dek">Codex CLI and Claude Code compared firsthand: sandbox and approval flags, AGENTS.md vs CLAUDE.md, real pricing, and where each terminal agent wins.</p>
+ <h3><a href="/blog/codex-cli-vs-claude-code/">Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026)</a></h3>
+ <p class="dek">Codex vs Claude Code, checked 29 Sep 2026: models, plans, usage limits, sandboxing, computer use, MCP and subagents in one dated table, then which to pick.</p>
<div class="foot">
<span class="tag kind">Comparisons</span>
- <a class="more" href="/blog/codex-cli-vs-claude-code/" aria-label="Read: Codex CLI vs Claude Code">Read →</a>
+ <a class="more" href="/blog/codex-cli-vs-claude-code/" aria-label="Read: Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026)">Read →</a>
</div>
</div>
</article>
diff --git a/docs/blog/tags/engines/index.html b/docs/blog/tags/engines/index.html
index a14432dc..b0d46341 100644
--- a/docs/blog/tags/engines/index.html
+++ b/docs/blog/tags/engines/index.html
@@ -179,13 +179,13 @@
<div class="meta">
<time datetime="2026-09-10">Sep 10, 2026</time>
<span class="dot" aria-hidden="true"></span>
- <span>5 min</span>
+ <span>7 min</span>
</div>
- <h3><a href="/blog/codex-cli-vs-claude-code/">Codex CLI vs Claude Code</a></h3>
- <p class="dek">Codex CLI and Claude Code compared firsthand: sandbox and approval flags, AGENTS.md vs CLAUDE.md, real pricing, and where each terminal agent wins.</p>
+ <h3><a href="/blog/codex-cli-vs-claude-code/">Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026)</a></h3>
+ <p class="dek">Codex vs Claude Code, checked 29 Sep 2026: models, plans, usage limits, sandboxing, computer use, MCP and subagents in one dated table, then which to pick.</p>
<div class="foot">
<span class="tag kind">Comparisons</span>
- <a class="more" href="/blog/codex-cli-vs-claude-code/" aria-label="Read: Codex CLI vs Claude Code">Read →</a>
+ <a class="more" href="/blog/codex-cli-vs-claude-code/" aria-label="Read: Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026)">Read →</a>
</div>
</div>
</article>
diff --git a/docs/blog/tags/multi-agent/index.html b/docs/blog/tags/multi-agent/index.html
index 2923d3a5..13dfd0e5 100644
--- a/docs/blog/tags/multi-agent/index.html
+++ b/docs/blog/tags/multi-agent/index.html
@@ -147,13 +147,13 @@
<div class="meta">
<time datetime="2026-09-10">Sep 10, 2026</time>
<span class="dot" aria-hidden="true"></span>
- <span>5 min</span>
+ <span>7 min</span>
</div>
- <h3><a href="/blog/codex-cli-vs-claude-code/">Codex CLI vs Claude Code</a></h3>
- <p class="dek">Codex CLI and Claude Code compared firsthand: sandbox and approval flags, AGENTS.md vs CLAUDE.md, real pricing, and where each terminal agent wins.</p>
+ <h3><a href="/blog/codex-cli-vs-claude-code/">Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026)</a></h3>
+ <p class="dek">Codex vs Claude Code, checked 29 Sep 2026: models, plans, usage limits, sandboxing, computer use, MCP and subagents in one dated table, then which to pick.</p>
<div class="foot">
<span class="tag kind">Comparisons</span>
- <a class="more" href="/blog/codex-cli-vs-claude-code/" aria-label="Read: Codex CLI vs Claude Code">Read →</a>
+ <a class="more" href="/blog/codex-cli-vs-claude-code/" aria-label="Read: Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026)">Read →</a>
</div>
</div>
</article>
diff --git a/docs/blog/topics/comparisons/index.html b/docs/blog/topics/comparisons/index.html
index 48ee66f1..3d3d28e0 100644
--- a/docs/blog/topics/comparisons/index.html
+++ b/docs/blog/topics/comparisons/index.html
@@ -317,13 +317,13 @@
<div class="meta">
<time datetime="2026-09-10">Sep 10, 2026</time>
<span class="dot" aria-hidden="true"></span>
- <span>5 min</span>
+ <span>7 min</span>
</div>
- <h3><a href="/blog/codex-cli-vs-claude-code/">Codex CLI vs Claude Code</a></h3>
- <p class="dek">Codex CLI and Claude Code compared firsthand: sandbox and approval flags, AGENTS.md vs CLAUDE.md, real pricing, and where each terminal agent wins.</p>
+ <h3><a href="/blog/codex-cli-vs-claude-code/">Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026)</a></h3>
+ <p class="dek">Codex vs Claude Code, checked 29 Sep 2026: models, plans, usage limits, sandboxing, computer use, MCP and subagents in one dated table, then which to pick.</p>
<div class="foot">
<span class="tag kind">Comparisons</span>
- <a class="more" href="/blog/codex-cli-vs-claude-code/" aria-label="Read: Codex CLI vs Claude Code">Read →</a>
+ <a class="more" href="/blog/codex-cli-vs-claude-code/" aria-label="Read: Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026)">Read →</a>
</div>
</div>
</article>
diff --git a/docs/blog/what-is-an-mcp-server/index.html b/docs/blog/what-is-an-mcp-server/index.html
index 9dd7e9ce..b1272b48 100644
--- a/docs/blog/what-is-an-mcp-server/index.html
+++ b/docs/blog/what-is-an-mcp-server/index.html
@@ -285,7 +285,7 @@ Commands:
</div>
<footer class="post-foot wrap"><div class="prevnext">
- <a class="pn prev" href="/blog/codex-cli-vs-claude-code/"><span class="k">← Older</span><span class="t">Codex CLI vs Claude Code</span></a>
+ <a class="pn prev" href="/blog/codex-cli-vs-claude-code/"><span class="k">← Older</span><span class="t">Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026)</span></a>
<a class="pn next" href="/blog/how-to-use-claude-code-plan-mode/"><span class="k">Newer →</span><span class="t">How to Use Claude Code Plan Mode</span></a>
</div>
<section class="related">
diff --git a/docs/llms.txt b/docs/llms.txt
index 6dddfad2..28455996 100644
--- a/docs/llms.txt
+++ b/docs/llms.txt
@@ -105,7 +105,7 @@
- [Aider vs Claude Code in 2026: Which Should You Use?](https://munderdiffl.in/blog/aider-vs-claude-code/): Aider vs Claude Code, checked 15 Sep 2026: how each edits code, which models it runs, what it costs, MCP, and Aider's stalled releases.
- [Claude Code vs Antigravity: Which One Fits Your Work?](https://munderdiffl.in/blog/claude-code-vs-antigravity/): Claude Code vs Google Antigravity, checked 14 Sep 2026: terminal agent or desktop agent app, Claude inside Antigravity, plans, limits and a verdict.
- [Claude Code vs Cursor: Which Is Better in 2026?](https://munderdiffl.in/blog/claude-code-vs-cursor/): Claude Code vs Cursor, checked 14 Sep 2026: where each runs, models, plans, the new Claude Code weekly limits, parallel agents and using both.
-- [Codex CLI vs Claude Code](https://munderdiffl.in/blog/codex-cli-vs-claude-code/): Codex CLI and Claude Code compared firsthand: sandbox and approval flags, AGENTS.md vs CLAUDE.md, real pricing, and where each terminal agent wins.
+- [Codex vs Claude Code: Plans, Limits and Computer Use (Sep 2026)](https://munderdiffl.in/blog/codex-cli-vs-claude-code/): Codex vs Claude Code, checked 29 Sep 2026: models, plans, usage limits, sandboxing, computer use, MCP and subagents in one dated table, then which to pick.
- [qm vs Munder Difflin: YC's Multiplayer Harness or an Office of Your Clones?](https://munderdiffl.in/blog/qm-vs-munder-difflin/): An honest September 2026 comparison of YC's qm and Munder Difflin. qm is a multiplayer agent harness a company deploys for Slack and the web. Munder Difflin runs an office of your clones on your own machine, with Teams to connect them.
- [CrewAI and AutoGen vs a Local Agent Harness: Framework or App?](https://munderdiffl.in/blog/crewai-autogen-vs-a-local-agent-harness/): CrewAI, AutoGen (now Microsoft Agent Framework), and LangGraph are frameworks — Python you write to build your own agent system. A multi-agent harness is an app you download that runs a team on your repos today. Here's how to tell which one you actually need.
- [GitHub Copilot CLI vs Claude Code as Hive Workers](https://munderdiffl.in/blog/github-copilot-cli-vs-claude-code-for-hive-work/): How GitHub Copilot CLI and Claude Code behave as agent engines inside a multi-agent harness: hooks, inbox mail, orchestration, session lifecycle — and why the right answer is to run both.
diff --git a/docs/sitemap.xml b/docs/sitemap.xml
index 2354beae..33589b6e 100644
--- a/docs/sitemap.xml
+++ b/docs/sitemap.xml
@@ -178,7 +178,7 @@
</url>
<url>
<loc>https://munderdiffl.in/blog/codex-cli-vs-claude-code/</loc>
- <lastmod>2026-09-10</lastmod>
+ <lastmod>2026-09-29</lastmod>
<changefreq>monthly</changefreq>
<priority>0.8</priority>
</url>