CSV Api
Upload a CSV. Get a REST API in seconds. No backend needed.
There is a new feature. Every dataset on your account can be turned into a MCP server that you can use with Claude, ChatGPT, and many other clients. This lets you talk with your datasets and get information from them without writing any code. All you do is add your dataset as a MCP and away you go!
Enjoy,
Sean
Added the ability to create a customizable shared view of your sheet. This sheet includes all of the available filters as well as a natural language input to describe your query. This has been great to share things with people quickly and easily.
Another area of improvement was the natural language query portion. Now you can get better search results using a natural language query than before. This can be seen in https://reedster.llc/ the projects section. This website is powered by csv-api and it's been working fantastic!
- Sean
Hey IH —
I just launched CSV-API (csv-api.com) and wanted to share it here first.
The problem I kept running into: I'd have a CSV — reference data, a client dataset, something for a prototype — and I needed to query it from an app. Every time I ended up writing the same throwaway Flask or Rails API, deploying it somewhere, and quietly maintaining it forever. Or just emailing the file around. Neither felt right.
So I built the tool I kept wishing existed.
Upload a CSV or Excel file, and you get a live REST API in under a minute. No backend, no deployment. Type inference runs automatically so your filters actually work. Every dataset gets interactive Swagger docs. There's a natural language query parameter if you'd rather write plain English than filter syntax.
It's free to start — 3 datasets, 500 rows, no credit card. Paid plans from $9/mo.
I'm a solo developer and this is a real product. I'd love two things from this community:
1. Try it on a dataset you actually have and tell me where it breaks or falls short
2. If the landing page isn't clicking, I want to know why
Happy to answer any questions about how it works or what's coming next.
-
1
Nice launch. The natural language query feature caught my eye — is that running on an LLM? If so, curious how you're managing API costs while keeping it free to start.
-
1
Hi, sorry for the late reply. I'm using ChatGPT's 5.4 nano model which is incredibly cheap and works fantastically for this use case. The biggest help is that the inputs are capped and the output is incredibly structured and small so the costs take a while to add up, a few dollars of api handles a lot of requests.
-
Comment