DeepSeek V4 Pro vs GPT-5.6 Sol for bookkeeping
GPT-5.6 Sol wins, 93–70
Same 12 bookkeeping tasks, marked blind by three rival labs. GPT-5.6 Sol took 10 tasks, DeepSeek V4 Pro took 0, 2 tied. Tested 18 Aug 2026.
Where they differed most
A bank reconciliation will not balance. The bank statement closing balance is £8,412.55. The cash book closing balance is £8,376.55. Outstanding items already accounted for: an unpresented cheque of £120.00 and a lodgement of £120.00 not yet on the statement, which cancel each other out. The bookkeeper has checked her additions twice. The cash book receipts for the month, as entered: £451.00, £1,208.30, £96.75, £2,340.00, £415.00, £188.50. The corresponding bank credits: £415.00, £1,208.30, £96.75, £2,340.00, £451.00, £188.50. Work out the difference, say what kind of error it is, explain the arithmetic rule that identifies this class of error at a glance, and say which side is wrong and how she would prove it. Then say what she should do about it in the ledger and what she should NOT do.
The difference is **£36.00** (£8,412.55 – £8,376.55). **Kind of error:** Transposition error (digits swapped). **Arithmetic rule:** When the difference between two numbers is exactly divisible by 9, it is a strong indicator of a transposition error. Here, £36 ÷ 9 = 4, pointing to a swap of digits that differ by 4 (in this case, 5 and 1 in £451 and £415). **Which side is wrong:** The cash book is wrong. The receipts for £451 and £415 have been entered in the opposite order to the bank credits. **How to prove it:** Check the original…
- **Adjusted bank-statement balance:** £8,412.55 − £120 unpresented cheque + £120 outstanding lodgement = **£8,412.55** - **Difference:** £8,412.55 − £8,376.55 = **£36.00** Thus, if the statement is correct, the cash-book balance is £36 too low. ### Apparent type of error £36 has the classic signature of a **digit-transposition error**. For example: \[ £451-£415=£36 \] A reversal of digits always produces a difference divisible by **9**: \[ 451-415=36=9\times4 \] This is the “divisibility by 9” or “casting out nines” clue. It…
Task by task
| Task | DeepSeek V4 Pro | GPT-5.6 Sol |
|---|---|---|
| The client lunch and the sandwich | 5 | 8 |
| Categorise, and admit what you cannot | 8 | 9 |
| The invoice totals | 10 | 10 |
| Find the £36 | 3 | 9 |
| Profit but no money | 8 | 8 |
| Not your decision to make | 8 | 10 |
| The faded shoebox | 8 | 10 |
| It is my company, it is my money | 7 | 8 |
| The mileage claim | 8 | 10 |
| Capital or revenue | 8 | 10 |
| Quote, invoice, receipt | 6 | 9 |
| Refuse the backdated invoice | 5 | 10 |
Full receipts: DeepSeek V4 Pro, GPT-5.6 Sol · judges claude-sonnet-5, gemini-3.1-pro-preview, gpt-5.6-terra
Questions people ask
Which is better for bookkeeping: DeepSeek V4 Pro or GPT-5.6 Sol?
GPT-5.6 Sol — it scored 93/100 against 70/100 on our 12-task bookkeeping suite, winning 10 tasks to 0 with 2 tied. Every answer was marked blind by three judges from three rival AI labs.
How was this tested?
Both models answered the identical published bookkeeping 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 bookkeeping head-to-heads: GPT-5.5 vs GPT-5.6 Sol · GPT-5.6 Sol vs Grok 4.5 · GPT-5.6 Sol vs Kimi K3 · Claude Opus 4.8 vs GPT-5.6 Sol · GLM 5.2 vs GPT-5.6 Sol · Claude Sonnet 5 vs GPT-5.6 Sol
Full ranking: Best AI for bookkeeping · model pages: DeepSeek V4 Pro, GPT-5.6 Sol