Video by Ian Nuttall (@iannuttall) · originally posted on XView original post on X
Claude Code (Claude Opus 4.1) vs OpenAI Codex (GPT-5) vs Cursor (GPT-5)
Next.js customer feedback app
Ian Nuttall@iannuttallHarness comparison1,218 likes195,403 views
Claude Code with Claude Opus 4.1, Codex with GPT-5 and Cursor's CLI agent with GPT-5 get the same prompt: a Next.js app with Tailwind 4 and shadcn that collects customer feedback and shows it in a widget, with 30 minutes each.
Each side pairs a coding harness with its own model, so harness and model differences are mixed. More Claude Code vs OpenAI Codex. All harness comparisons.
Harnesses and models
| Harness | Model | Tokens | Notes |
|---|---|---|---|
| Claude Code | Claude Opus 4.1Anthropic | 33K tokens | Working app with no errors after one correction; basic design; fastest.Written as “Opus 4.1” |
| OpenAI Codex | GPT-5OpenAI | 102K tokens | No working app after 30 minutes (stuck on Tailwind 4 setup, “failed to compile”). |
| CursorCursor Agent (CLI) | GPT-5OpenAI | 188K tokens | Slowest; working app after one or two small fixes; minimal design. |
Cost & time as stated in the post: Each harness got the same prompt and 30 minutes. Token counts are as the post states them.
Details
| Kind | Side-by-side (same post, same prompt) |
|---|---|
| Task | Next.js customer feedback app |
| Author’s verdict | “Opus is still the more capable model out of the box and Claude Code is the more complete CLI product” — @iannuttall’s view |
| Category | Web/UI |
| Mode | Not stated |
| Posted | |
| Video length | 78.3 s |
| Likes / views | 1,218 / 195,403 · as of Oct 18, 2026 |
| Bookmarks | 835 |
| Sponsored | No |
| Original post | x.com/iannuttall/status/1953833034794651649 |
Related
Content belongs to its creators. Video by @iannuttall; aivs.tv hosts a copy for fast playback and links to the original post on X. Request removal
