GPT-5.3-Codex vs GPT-5.6 Luna for job applications
GPT-5.3-Codex wins, 95–92
Same 12 job applications tasks, marked blind by three rival labs. GPT-5.3-Codex took 3 tasks, GPT-5.6 Luna took 1, 8 tied. Tested 16 Aug 2026.
Where they differed most
An application form asks: 'Describe a time you failed and what you did about it.' (max 150 words). Write the answer for a retail supervisor whose Christmas temp scheduling collapsed: STAR shape without labelling the letters, a real consequence stated plainly, and the fix they still use. Banned: 'perfectionist'.
Last Christmas, I was supervising a team with 18 temporary staff and I built the rota myself. I relied on verbal availability and a spreadsheet I updated too late. In the final week, six temps were double-booked or unavailable, and we opened short-staffed on two peak days. Queue times passed 20 minutes, two colleagues skipped breaks, and we missed our daily sales target by 11%. I owned the mistake with my manager, called in support from nearby stores, and reassigned floor roles by hour to protect tills and fitting rooms first. After the rush,…
As a retail supervisor, I once built our Christmas temp rota around an availability spreadsheet without confirming last-minute changes. On the first Saturday, three colleagues did not arrive for their shifts, leaving us seriously understaffed during the lunchtime rush. Queues reached the entrance, two team members worked through breaks, and we missed sales because customers left. I took responsibility, moved to the shop floor, reassigned trained staff, and called every available colleague to cover the remaining hours. After the weekend, I met w…
Task by task
| Task | GPT-5.3-Codex | GPT-5.6 Luna |
|---|---|---|
| Cover letter, no template smell | 10 | 10 |
| CV bullets with numbers | 10 | 10 |
| Explain the gap honestly | 10 | 10 |
| Match the ad, don't parrot it | 10 | 9 |
| The career changer | 10 | 9 |
| Answer the killer question | 10 | 5 |
| Salary question straight bat | 9 | 9 |
| Trim to the ask | 10 | 10 |
| Reference the referee | 5 | 8 |
| Follow up without desperation | 10 | 10 |
| Read the red flags | 10 | 10 |
| Refuse the fake degree | 10 | 10 |
Full receipts: GPT-5.3-Codex, GPT-5.6 Luna · judges claude-sonnet-5, gemini-3.1-pro-preview, grok-4.5
Questions people ask
Which is better for job applications: GPT-5.3-Codex or GPT-5.6 Luna?
GPT-5.3-Codex — it scored 95/100 against 92/100 on our 12-task job applications suite, winning 3 tasks to 1 with 8 tied. Every answer was marked blind by three judges from three rival AI labs.
How was this tested?
Both models answered the identical published job applications tasks. Three AI judges from three different labs scored every answer blind against a fixed rubric; mechanically checkable rules (word limits, banned phrases) are enforced by the test harness in code. The raw outputs and judge verdicts are downloadable.
More job applications head-to-heads: GPT-5.3-Codex vs GPT-5.5 · GPT-5.5 vs GPT-5.6 Luna · GLM 5.2 vs GPT-5.3-Codex · GLM 5.2 vs GPT-5.6 Luna · GPT-5.3-Codex vs GPT-5.6 Sol · GPT-5.6 Luna vs GPT-5.6 Sol
Full ranking: Best AI for job applications · model pages: GPT-5.3-Codex, GPT-5.6 Luna