Setu and SpendixAI are not alternatives, because they have different customers. Setu is positioned as a paid library and API provider with Indian fintech APIs, including Account Aggregator, so its customers are builders. This tracker is a finished expense tracker for individuals in India. If you are building a product, look at the first. If you want to see your spending, look at the second.
How do the two differ by customer?
| Question | Setu (positioned as) | This tracker |
|---|---|---|
| Who uses it? | Developers and fintech teams | Individuals tracking personal spending |
| What do you receive? | APIs, per its positioning | A finished app on web and mobile |
| Pricing | Paid; none verified here | Free; Pro ₹99 monthly, ₹999 yearly, ₹24,999 once |
| Data route | Not detailed in our sources | Statements as PDF, CSV or Excel, plus typed entries |
| Bank connection | Account Aggregator is in its positioning | None; no bank sync, no SMS reading |
Details of Setu's products belong to its own pages. We do not describe them here.
Could a startup build its own tracker on an API instead?
Yes, and many do. Imagine a team that needs 10,000 users' transactions each month. If a provider charged a made-up ₹2 per user per month, the cost is 10,000 x 2 = ₹20,000 a month, before the team's own engineering, support and compliance work. A ready product costs the individual ₹99 a month. Neither number is a real quote; the point is that building and buying carry different shapes of cost.
What does the end-user route look like?
- Sign in with Google; there are no passwords.
- Choose the account for the file you are importing.
- Upload one or several statements.
- Review the rows, fix categories and save.
- Open reports or the Report builder, or download CSV.
The same workspace opens on any device. Records live in your own workspace with role-based access, and owners can export data, choose document retention and delete transactions.
Is the MCP connection an API?
Not in the developer sense. You can connect Claude or ChatGPT through MCP, choose a scope, start read-only and disconnect any time, which lets an assistant answer questions from your own records. It is a user feature, not a way to build a product on top of this tracker.
When this tracker is not the right fit
If you need developer APIs, consent flows to embed in your own product, or direct institutional data, this tracker is not it. It also has no team workspace and no bank sync.
What should a founder weigh before building instead of buying?
Weigh time as well as money. Building means consent screens, categorisation, error handling and a support inbox, all before a single user is happy. If your edge is elsewhere, an existing tracker may serve your own needs while you build the thing you are really good at. If your edge is exactly this, an API provider is the natural starting point, and the comparison table from the vendor's own documentation is the next stop.
For a user rather than a builder, the same logic holds in miniature. An upload route needs a few minutes each month and never asks you for a banking login. If that rhythm suits you, there is nothing to integrate and nothing to build. If you would rather have an automatic feed, say so clearly, because that preference is the one thing this tracker does not meet.
In short
- Builders: weigh providers against written requirements.
- Individuals: test a tracker with a real statement.
- Everyone: check consent, export and deletion first.
Useful ideas start with a clearer picture of your own spending.
Explore your reports


