Free · No sign-up · No login

Find out what an SIH judge would say about your idea.

Paste your idea below. You'll get a score out of 10 across the seven criteria Smart India Hackathon judges actually weigh — plus the specific gaps they'd push back on, and what to fix first.

Add a one-line title for the problem.
0 characters · minimum 200 — the more specific you are, the more useful the score.
Please describe your idea in a little more detail.

We store ideas anonymously to improve the scoring. No names, no emails, nothing that identifies you — and we never share individual submissions.

Scoring your idea

This usually takes a minute or two.

0:00
  1. Received and queued
  2. Reading your write-up
  3. Scoring against the seven criteria
  4. Writing up gaps and next steps
  5. Ready
Your result link — you can close this page and come back

Your scorecard

Permanent link to this scorecard — valid for 7 days
The rubric

Seven factors, weighted

Your overall score is a weighted average — not a number the model picks on a whim. Novelty and impact carry the most weight, because that's what separates a shortlisted idea from a competent one.

Before you submit

Why most SIH ideas score badly

Four patterns account for most low scores. None of them are about the idea being bad.

Most common

The write-up describes a category, not a solution

"An AI-powered platform for farmers" tells a judge nothing. Which farmers, which decision, which data, what does the model actually predict? Specificity is the single cheapest way to raise your score.

Kills novelty

It already exists and you didn't say why yours differs

Judges have seen forty crop-disease detectors. If yours works offline, or on a ₹500 device, or on a crop nobody covers — lead with that. If you can't name the difference, you have a positioning problem, not a writing problem.

Kills depth

Every hard part is hidden behind an API

Calling a vision API is not computer vision. Judges look for the part you actually engineered — the sync protocol, the edge deployment, the data pipeline, the constraint you designed around. Name it explicitly.

Kills feasibility

It cannot be built in 36 hours

A national-scale integrated platform is not a hackathon deliverable. Scope to a prototype that demonstrates the hard part, and say what the prototype proves. Ambition in the vision, realism in the build.

Questions

Frequently asked

No. This is an independent tool, not affiliated with or endorsed by the Smart India Hackathon or the Ministry of Education. The rubric reflects the criteria SIH publishes and that evaluators commonly apply, but the score has no official standing whatsoever. Treat it as a rehearsal, not a verdict.
Usually it means your description is thin, not your idea. The model can only judge what you wrote. Most teams who re-score after adding technical detail jump one to two points without changing the idea at all. Read the Gaps column — that's the useful part.
Ideas are stored anonymously for improving the tool, and individual submissions are never published or shared. That said — it is sent over the internet to an AI model, so apply normal judgement. If some part of your idea is genuinely patentable, describe it in general terms rather than pasting the specifics.
Requests are queued so that everyone gets scored even when a lot of teams submit at once — a queue is what stops the tool falling over during the rush before a deadline. You get a permanent link the moment you submit, so you can close the tab and come back to it. If the queue is long, your result simply arrives a bit later; the link stays valid for seven days.
No — you get the same one. Results are keyed to the text of the idea, so resubmitting unchanged text just returns the result you already have. Change the write-up and you get a fresh score, which is exactly how you're meant to use it.
As often as you like, within reason — there's a short pause between runs and a cap of about a dozen per ten minutes to keep the tool free for everyone. Rewriting and re-scoring is exactly how you're meant to use it.
No. Real judging includes your presentation, your prototype, your team, and how you handle questions — none of which this sees. A high score means your written idea holds up to scrutiny. That's a necessary condition, not a sufficient one.

Rewrite it. Re-score it. Then submit it.

Most teams gain a point or two on the second pass. It takes five minutes.

Score My Idea