Free Patent Health Check: How to Score a Patent

This article explains how the points are awarded, using the rules published on the Patent Checkup page (rules version 6), so that you can read a score critically, or estimate one yourself on paper.

What you put in, and what happens to it

You can enter a patent number (including the number from a Chinese certificate starting with ZL, or a publication number from the US, Europe, China and other offices), or upload the patent PDF. The page says that:

• A lookup by number reads the public patent page.
• An upload uses only the patent number or claims found in the file.
• Uploaded files are not stored, not written to logs and not sent with the contact form.
• The checkup is free and needs no account. Your details reach the firm only if you choose to have a lawyer look at the patent and send the form.

The Chinese-language site works differently: it does not offer lookup by number, and you upload the granted patent document instead.

The three parts of the score

1. Scope of protection: up to 5 points

This asks how broadly claim 1 is drafted. Fewer and shorter limitations mean a wider scope. Two measures:

Number of limitations in claim 1 (up to 3 points)

Limitations	Points
Under 4	3
4 to under 6	3
6 to under 8	2
8 to under 10	1
10 to under 13	1
13 or more	0

Length of claim 1 (up to 2 points; English claims are counted in characters, with different bands from Chinese)

English claim length (characters)	Points
Under 501	2
501 to under 1,101	1
1,101 to under 1,501	1
1,501 or more	0

Chinese claims use their own bands, counted by characters: under 151 earns 2 points, 151 to under 351 earns 1, 351 to under 501 earns 1, and 501 or more earns 0.

2. Fallback positions: up to 5 points

This asks how many positions are left if claim 1 is challenged: the number of dependent claims, and whether the independent claims cover more than one category (product, method, use).

Dependent claims	Points
Under 3	0
3 to under 10	1
10 to under 15	2
15 to under 20	2
20 or more	3

Categories covered by independent claims	Points
Under 2	0
2 or more	2

This part needs the complete, numbered claims to count the dependent claims. If they cannot be read, this part is not scored.

3. Design-around difficulty: up to 90 points

This is where almost all the points are. The checkup takes the features of claim 1 and asks, for each one, whether another way could serve the same purpose, or whether the feature could be left out. The more such features, the lower the score.

Two AI models from different vendors each read claim 1 on their own, without seeing the other's answer. Each gives every feature one of three judgments:

• Hard to avoid: no other way serves the same purpose, and the feature cannot be left out.
• Removable: the product still serves the same purpose without this feature. Counts as bypassable.
• Replaceable: another way, outside the literal wording of the claim, serves the same purpose. Counts as bypassable.

The checkup then makes a weighted count of the features flagged as replaceable or removable. A feature flagged by both models counts as 1. A feature flagged by only one model counts as 0.5. The AI does not award points: the published bands below convert the weighted count into points. The page also shows a percentage of features flagged, but that figure is for display only and does not affect the score.

Weighted count of replaceable or removable features	Points
0	90
0.5	80
1	70
1.5	60
2 to under 3	50
3 to under 4	40
4 to under 5	30
5 to under 6	20
6 to under 7	10
7 or more	0

Without this AI step there is no total at all. If it does not complete, the page says to run the checkup again a little later, which usually works.

The grade

Total	Grade
72 and above	Solid
58 to 71	Good
45 to 57	Fair
Below 45	Needs work

If one part cannot be read, it is not scored and the total is scaled to 100 over the parts that were. When you check by patent number, the judgments are archived, so the same patent gets the same score again.

A worked example (invented)

Take an invented patent for a battery cooling plate. I made up all the numbers.

• Claim 1 has 6 limitations and 700 English characters → 2 + 1 = 3 points for scope.
• The patent has 12 dependent claims, and independent claims for a product and a method → 2 + 2 = 4 points for fallback.
• Both models flag the same one feature of claim 1 as replaceable, and neither flags any other. The weighted count is 1 (a feature flagged by both models counts as 1, and there are no half-counts here) → 70 points for design-around.

Total: 3 + 4 + 70 = 77, "Solid".

Now a contrasting invented patent for a clamp assembly:

• Claim 1 has 11 limitations and 1,600 English characters → 1 + 0 = 1 point for scope.
• Only 2 dependent claims and a single category → 0 + 0 = 0 points for fallback.
• Model A flags 5 features of claim 1 and model B flags 5. Four of them are flagged by both, and each model flags one more that the other does not. The weighted count is 4 × 1 + 2 × 0.5 = 5 → band "5 to under 6" → 20 points for design-around.

Total: 1 + 0 + 20 = 21, "Needs work".

The lesson from both: because design-around difficulty carries 90 of 100 points, the grade mostly reflects how many features of claim 1 a competitor could replace or drop, and whether both models agree. Short, many-claim patents do not automatically score well.

What the score is useful for

• Triage across a list of patents. If you inherited a portfolio or are looking at patents to buy or license, a consistent first pass helps decide which ones deserve a lawyer's time. For the money side of that decision, read Patent Maintenance Cost Analysis: When to Abandon Low-Value Patents.
• Reading a competitor's patent. The design-around part tells you which features of claim 1 look replaceable. My article The Weakest Limitation shows how to turn that into design options, and the Design-Around tool is built for this.
• Spotting a thin patent early. A patent with few dependent claims and one category has little to fall back on if claim 1 is challenged. See Patent Invalidity Defense for how challenges work.

What the score does not tell you

The page is direct about this, and so should this article be:

• It is not a conclusion about legal effect. The score reflects structural features that can be read from the public document, plus two AI models' judgments of claim 1, feature by feature. Whether the scope would be upheld depends on the product and the prior art, and a lawyer judges that.
• A high score does not mean the patent is safe to rely on.
• It looks at claim 1 only for scope and design-around. Other independent claims are used only to see whether more than one category is covered.
• It does not search for prior art, check legal status or compare with your product. For that, use the FTO Screen for product-versus-patent questions, and a professional for anything that matters.

If you want to know what a claim actually protects before you read a score, start with What Are Patent Claims?.

Try it

Run the free Patent Checkup with a patent number, or upload the PDF, and compare its output to the tables above. Each part of the score is explained on the results page. If you would like a lawyer to look at a patent afterwards, the page offers that as an optional step.

General information only, not legal advice. Using our tools does not create an attorney-client relationship. For advice on your situation, consult a registered patent attorney or agent.