+ Follows the requested order and consistently gives actions and stopping costs.
- Treats coffee and convenience-store purchases as merchant repeats and infers some losses.
Catches duplicate services and convenience premiums, and never tells you to stop any of it.
| Category | Money › Budgeting |
|---|---|
| Tags | AnalyzingReviewingChecklist |
**Open the output with this, written in the language you are answering in: "This is not financial advice and does not replace a professional."** ***Say it in the reader's language, not in English*** — **a disclaimer the reader skims past does nothing.** Find where money is going that I have not noticed. **Look for, in this order:** 1. **Recurring charges I did not list as recurring** — ***the same merchant appearing more than twice*** 2. **Things paying for the same need twice** — overlapping services, two subscriptions covering one thing 3. **Convenience premiums** — **the same item bought in a more expensive way.** *Delivery over pickup, small pack over large, late over planned* 4. **Charges tied to a decision I already reversed** — a service for something I stopped doing 5. **Fees** — transfer, late, foreign currency, minimum-balance, ATM. ***These are small and pure loss*** 6. **Free trials that converted** 7. **Spending that clusters** — several purchases on the same day or after the same trigger. **Say what the pattern is, not what it means about me** **Per finding:** what it is / how often / **what action would stop it, concretely** / ***whether stopping it costs something*** ⚠️ ***Do not calculate or total anything.*** **Do not estimate what I would save.** **Do not tell me to stop any of it** — some of these are worth paying for and I am the one who knows which. **Where the data is not enough to tell, say so rather than inferring a pattern.**
Leaks are a problem because they repeat, and duplicates or convenience premiums are invisible line by line. This finds the pattern and leaves the decision to you — some of it is worth paying for.
Gemini leads on compliance and completeness. ChatGPT is the most concise but weak on same-merchant evidence, while Claude is specific yet makes more unsupported inferences.
+ Follows the requested order and consistently gives actions and stopping costs.
- Treats coffee and convenience-store purchases as merchant repeats and infers some losses.
+ Its concise structure and restraint around missing data are especially strong.
- Classifies repeated categories without establishing that the merchant was the same.
+ Provides highly specific next steps and identifies the data needed for verification.
- Assumes a monthly gym charge and infers a coffee price premium without evidence.
| Criterion | ChatGPT | Claude | Gemini | Leader |
|---|---|---|---|---|
| Instruction following | 8 | 8 | 9 | Gemini +13% |
| Accuracy | 8 | 7 | 8 | Tie |
| Specificity | 8 | 9 | 9 | Tie |
| Structure | 9 | 9 | 9 | Tie |
| Right length | 9 | 8 | 8 | ChatGPT +13% |
Scored 1–10 by gpt-5.6-sol with model names hidden (2026-09-24). This is an AI review, not a measurement.
We gave three models the same input and copied their answers unedited. Each ran in its CLI (an agent harness), and answers in the ChatGPT or Claude apps or on the web may differ. Outputs are in Korean.
My spending: 스트리밍A 13,900 (매월) / 스트리밍B 9,900 (매월) / 음악 8,900 (매월) / 클라우드 2,900 (매월) / 배달 대행료 3,000×11회 / 편의점 소액 27건 / 커피 4,500×18회 / 헬스장 69,000 (3개월째 미방문) / 해외결제 수수료 1,200×4 / 무료체험 전환 5,900
이 내용은 금융 조언이 아니며 전문가의 조언을 대신하지 않습니다.
배달 대행료
커피
편의점 소액 구매
스트리밍A와 스트리밍B
스트리밍 서비스와 음악 서비스
해외결제 수수료
그 밖의 수수료
| Prepare a money conversation with a partner | |
| Work backwards from a money goal | |
| Build a budget from what you actually spend | |
| Re-sort spending by what could have been different | |
| Audit your subscriptions |