Best AI for everyday maths / head-to-head
GPT-5.6 Luna vs GPT-5.6 Terrafor everyday maths
GPT-5.6 Terra wins, 99–98
Same 12 everyday maths tasks, marked blind by three rival labs. GPT-5.6 Terra took 1 tasks, GPT-5.6 Luna took 0, 11 tied. Tested 14 Aug 2026.
Where they differed most
Roughly how many working hours are there in a UK full-time year? Give your estimate, show the 2-3 assumptions it rests on, and say which assumption moves the answer most.
**Rough estimate: about 1,750 actual working hours per year** for a UK full-time employee.
Assumptions:
1. **37.5 contracted hours per week**.
2. **5.6 weeks’ annual leave**, the UK statutory minimum, including bank holidays.
3. **No substantial sick leave, unpaid leave, or overtime**.
Calculation:
\[
(52 - 5.6)\times 37.5 \approx 1,740\text{ hours}
\]
So a sensible range is **about 1,700–1,850 hours**:
- 37.5 hours/week with typical leave: roughly **1,700–1,750**
- 40 hours/week with statutory leave: roughly **1,856**
The assumption tha…A reasonable estimate is **about 1,700–1,750 working hours per year** for a UK full-time employee. Example estimate: **1,740 hours/year** Assumptions: 1. **37.5 hours per week** (e.g. 9–5 with a 30-minute unpaid lunch) - 37.5 × 52 = **1,950 hours** 2. **28 days’ paid annual leave**, including bank holidays — the UK statutory minimum for a five-day worker - 28 days × 7.5 hours = **210 hours** 3. No allowance for sickness absence, unpaid leave, or other non-working time - 1,950 − 210 = **1,740 hours** The assumption that move…
Task by task
| Task | GPT-5.6 Luna | GPT-5.6 Terra |
|---|---|---|
| VAT both ways | 10 | 10 |
| Percentage change vs points | 10 | 10 |
| Unit price comparison | 10 | 10 |
| Mortgage overpayment intuition | 10 | 10 |
| Split a bill with a twist | 10 | 10 |
| Compound vs simple | 10 | 10 |
| The discount stack | 10 | 10 |
| Reverse percentage | 10 | 10 |
| Rate to reality | 9 | 9 |
| Average of averages trap | 10 | 10 |
| Estimate honestly | 9 | 10 |
| Refuse false precision | 10 | 10 |
Full receipts: GPT-5.6 Luna, GPT-5.6 Terra · judges claude-sonnet-5, gemini-3.1-pro-preview, grok-4.5
Questions people ask
Which is better for everyday maths: GPT-5.6 Luna or GPT-5.6 Terra?
GPT-5.6 Terra — it scored 99/100 against 98/100 on our 12-task everyday maths 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 everyday maths 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 everyday maths head-to-heads: Claude Opus 4.8 vs GPT-5.6 Terra · Claude Opus 4.8 vs GPT-5.6 Luna · GPT-5.6 Sol vs GPT-5.6 Terra · GPT-5.6 Luna vs GPT-5.6 Sol · Claude Opus 4.6 vs GPT-5.6 Terra · GPT-5.3-Codex vs GPT-5.6 Terra
Full ranking: Best AI for everyday maths · model pages: GPT-5.6 Luna, GPT-5.6 Terra