GPT-5.3-Codex vs GPT-5.6 Sol for humanising AI text
GPT-5.3-Codex wins, 97–94
Same 12 humanising AI text tasks, marked blind by three rival labs. GPT-5.3-Codex took 1 task, GPT-5.6 Sol took 0, 11 tied. Tested 16 Aug 2026.
Where they differed most
Rewrite this product description naturally, keeping every spec — 750ml, stainless steel, cold 24 hours, hot 12, £24, free engraving: 'Introducing the ultimate hydration solution for the modern lifestyle. This premium 750ml vessel is expertly crafted from high-grade stainless steel, delivering unparalleled temperature retention — beverages remain refreshingly cold for an impressive 24 hours or delightfully hot for 12. Elevate your everyday experience for just £24, with complimentary personalised engraving included.' Sound like a good market stall, not a press release.
Right then — a proper **750ml stainless steel bottle** here, built to last. Keeps your drink **cold for 24 hours** or **hot for 12**, no bother. Only **£24**, and I’ll do **free engraving** on it as well — perfect for gifts or keeping yours from going walkabout.
Stay refreshed all day with this 750ml stainless steel bottle. It keeps drinks cold for up to 24 hours or hot for up to 12 hours. Just £24, including free personalised engraving.
Task by task
| Task | GPT-5.3-Codex | GPT-5.6 Sol |
|---|---|---|
| The touching-base email | 10 | 10 |
| De-guru the bio | 10 | 10 |
| Product page, human voice | 10 | 7 |
| Strip the tics | 9 | 9 |
| The head's newsletter | 10 | 10 |
| The condolence card | 10 | 10 |
| Reply to the review, humanly | 10 | 10 |
| About us, actually about us | 10 | 10 |
| Best man, not best template | 9 | 9 |
| Human but still formal | 9 | 9 |
| Change log required | 9 | 9 |
| Refuse the disguise job | 10 | 10 |
Full receipts: GPT-5.3-Codex, GPT-5.6 Sol · judges claude-sonnet-5, gemini-3.1-pro-preview, grok-4.5
Questions people ask
Which is better for humanising AI text: GPT-5.3-Codex or GPT-5.6 Sol?
GPT-5.3-Codex — it scored 97/100 against 94/100 on our 12-task humanising AI text suite, winning 1 tasks to 0 with 11 tied. Every answer was marked blind by three judges from three rival AI labs.
How was this tested?
Both models answered the identical published humanising AI text 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 humanising AI text head-to-heads: Claude Sonnet 5 vs GPT-5.3-Codex · Claude Fable 5 vs GPT-5.3-Codex · GPT-5.3-Codex vs GPT-5.5 · GLM 5.2 vs GPT-5.3-Codex · GPT-5.3-Codex vs GPT-5.6 Terra · GPT-5.3-Codex vs Qwen3.7 Max
Full ranking: Best AI for humanising AI text · model pages: GPT-5.3-Codex, GPT-5.6 Sol