Patent search cost varies by a factor of forty across search types, and the reason is not effort. It is the standard each search has to meet.
A patentability search can stop when it finds one good reference. If something already discloses your invention, that answers the question.
A freedom-to-operate search cannot stop. It has to find every in-force claim your product might infringe, in every market you sell into, and missing one is the failure mode.
That difference is most of the price gap.
Why the ranges are so wide
Two quotes for "a patent search" can differ fivefold without either being unreasonable.
| Variable | Cheap end | Expensive end |
|---|---|---|
| Jurisdictions | US only | US, EP, JP, CN and more |
| Literature | Patents only | Plus journals, conferences, standards, manuals |
| Languages | English | Multiple, with specialist readers |
| Depth | Screening | Exhaustive, litigation-grade |
| Deliverable | Reference list | Element mapping and analysis |
| Turnaround | Standard | Rush |
Always compare the deliverable, not the headline number. A search producing a list of twenty references is a different product from one mapping each reference against your intended claim elements.
What each type costs
| Search type | Typical cost | Answers |
|---|---|---|
| Self-search, free tools | $0 | Is this obviously described already? |
| Patentability | $500–$3,000 | Can I get a patent on this? |
| Freedom to operate | $3,000–$15,000+ | Can I sell this without infringing? |
| Validity | $5,000–$20,000+ | Can this patent be invalidated? |
| Landscape | Priced per project | Who holds what in this area? |
| Attorney opinion on results | Additional | What do these references mean? |
The opinion is separate from the search. A search firm produces references; an attorney interprets them. Some engagements bundle both, and the distinction matters when comparing quotes.
Ranges are wide because scope varies. Two "freedom to operate" quotes can differ threefold on jurisdiction count and claim element coverage alone.
What drives the price
| Factor | Effect |
|---|---|
| Number of jurisdictions | Each adds a separate body of art |
| Non-patent literature included? | Journals, conferences, manuals, standards — expensive to cover |
| Foreign-language sources | Translation and specialist reading |
| Claim elements traced | FTO must map every element of every relevant claim |
| Technology complexity | Biotech and chemistry take longer than mechanical |
| Turnaround | Rush service carries a premium |
| Written opinion | Legal work on top of search work |
Non-patent literature is the biggest single cost driver and the most often omitted from cheap quotes. It is also where the references that defeat you frequently hide.
Ask what is excluded, not just what is included. A quote that covers US patents only, without non-patent literature, is a different product from one that does not.
The hidden cost of not searching
The expense is rarely the search. It is what follows from skipping it.
| Skipped at | Discovered at | Cost of the gap |
|---|---|---|
| Idea stage | Provisional | ~$2,000 |
| Provisional | First office action | $12,000+ and 2 years |
| Filing | Grant | The full spend |
| Grant | Litigation | The patent and the case |
Every row down that table multiplies the cost of the same information. The reference was findable throughout; only the price of finding it changed.
When free searching is enough
| Situation | Free search sufficient? |
|---|---|
| Curiosity about a product | Yes |
| First check on an idea | Yes |
| Before a provisional | Yes, as a minimum |
| Before a non-provisional | No |
| Before a product launch | No |
| Defending an infringement claim | No |
| Portfolio renewal triage | Usually yes |
Free tools eliminate most inventions, which is exactly what you want them to do. Most ideas have been described before, and finding that out for nothing is the best available outcome.
The threshold is where real money starts. Once you are committing $8,000 to $20,000 to a non-provisional, $1,500 to check the ground is proportionate.
See how to see if something is patented for the free method in full.
Worked example: cost of searching versus cost of not
An inventor with a mechanical improvement.
| Path | Search spend | Outcome | Total spent |
|---|---|---|---|
| A: search first | $0 free + $1,800 professional | Anticipated — stops here | $1,800 |
| B: file first | $0 | Provisional, then non-provisional, then §102 rejection at month 27 | ~$14,000 |
| C: no search at all | $0 | Patent granted on narrow claims; invalidated in year 6 | $20,000+ and the case |
What path B actually cost
| Item | Cost |
|---|---|
| Provisional | $2,100 |
| Non-provisional | $9,300 |
| First office action response | $2,600 |
| Elapsed time | 27 months |
| Outcome | Abandoned |
The reference the examiner found was in the free databases the whole time. Path A found it in an afternoon.
Path C is the expensive one and the least visible. A patent that issues over missed prior art looks like a success for years, and fails at the moment it matters — when someone with a litigation budget looks properly.
Cheaper ways to get more from the same budget
| Tactic | Effect |
|---|---|
| Free search first, share the results | Searcher's hours go to new ground |
| Narrow the technology statement | Fewer wasted hours on adjacent art |
| Name the field's synonyms | Crosses the vocabulary gap immediately |
| Specify jurisdictions tightly | Do not pay for markets you will not enter |
| Order search without opinion initially | Add the opinion only if results warrant |
| Batch related inventions | Some overlap in the searching |
Sharing your own results is the single biggest lever. A searcher starting from zero repeats work you have done; one starting from your references spends the budget on ground you could not reach.
Ordering the search separately from the opinion lets you stop after the search if it comes back clearly negative, which happens often enough to matter.
Search cost against filing cost
| Cost | |
|---|---|
| Free self-search | $0 |
| Patentability search | $500–$3,000 |
| Non-provisional to grant | $8,000–$20,000 |
| Maintenance fees, large entity | $14,470 |
| Lifetime total | $22,000–$34,000+ |
The search is roughly 5–10% of the lifetime cost of the patent it protects, and it is the only part that can prevent the rest being wasted.
Small entities pay 40% of USPTO fees and micro entities 20%, which lowers the fee side but not the search or attorney side. See how much does a patent cost.
Searching at renewal, not just at filing
The cheapest useful search is often the one nobody runs.
| Ipiry Patent Survival Curve v1.0 | Rate |
|---|---|
| Survive the 3.5-year fee (2022 cohort) | 85.8% |
| Survive the 7.5-year fee (2018 cohort) | 64.6% |
| Reach full term (2014 cohort) | 41.4% |
| Abandoned before full term | 58.6% |
| Lost at the third fee | 23.2 points |
Computed from 27,273,654 USPTO maintenance fee records covering 8,262,336 US utility patents — see the patent survival curve.
The third maintenance fee is $8,280 for a large entity. A search establishing whether any competitor practises the claims costs a fraction of that and turns the decision from instinct into evidence.
Most abandonments happen without that check, because it historically cost more than the fee. That is no longer true, and it is the strongest argument for searching at renewal. See patent portfolio management.
What you get for each price band
| Band | Typically includes | Typically excludes |
|---|---|---|
| $0 | Patent full text, families, status | Non-patent literature, analysis |
| $500–$1,000 | US patents and applications, basic report | Foreign art, NPL, opinion |
| $1,500–$3,000 | Multi-jurisdiction, non-patent literature, element mapping | Legal opinion |
| $3,000–$15,000 | FTO scope, every in-force claim, multiple markets | — |
| $5,000+ | Validity depth, exhaustive, litigation-grade | — |
The jump from $1,000 to $1,500 usually buys non-patent literature, and that is the most valuable increment in the table. Conference papers and trade publications are where damaging references hide.
Below $500 the search is usually US patents only. That is a useful first pass and it is not a basis for a filing decision on anything commercially significant.
Free searching is not zero cost
It costs time, and time is the thing people are usually short of.
| Task | Realistic time |
|---|---|
| First keyword pass | 1 hour |
| Semantic search and triage | 1 hour |
| Classification browsing | 1–2 hours |
| Reading close references properly | 2–4 hours |
| Market and product searching | 1 hour |
Reading is the majority of it. Finding fifty documents takes minutes; deciding which of them matter is the work, and skipping it makes the search worthless.
Which reframes the cost comparison. A $1,800 professional search against six hours of your own time is a straightforward calculation once you price the hours honestly.
Getting a useful quote
| Ask | Why |
|---|---|
| Which jurisdictions? | Biggest scope variable |
| Is non-patent literature included? | Biggest cost variable |
| Foreign-language coverage? | Often excluded quietly |
| What does the deliverable contain? | References only, or analysis |
| Is an opinion included? | Usually separate |
| Turnaround? | Rush costs more |
| What is explicitly out of scope? | The most revealing question |
Compare deliverables, not prices. A $700 search and a $2,500 search are frequently different products rather than different rates for the same work.
Do the free search first and share it. A searcher who starts with what you have already found spends their hours on the areas you could not reach.
Who pays, and when
| Situation | Who typically bears it |
|---|---|
| Pre-filing search for a client | Billed to the client |
| In-house R&D screening | R&D or IP budget |
| Due diligence on an acquisition | The buyer |
| Validity search when sued | Litigation budget |
| Renewal triage | IP budget, against the fee saved |
Renewal triage is the case where the search pays for itself directly. A search costing a few hundred dollars against an $8,280 third maintenance fee is self-funding if it releases even one patent that would otherwise have been renewed by default.
Budget searching as a line item, not an exception. Practices that treat it as optional tend to skip it precisely when the stakes justify it.
Patent search cost: the checklist
- Always run the free search first. It costs an hour and eliminates most inventions.
- Search products and market listings too, since a product on sale is prior art whether or not anyone patented it.
- Match the search type to the decision — patentability before filing, FTO before launch, validity when challenged.
- Commission a professional search before the non-provisional, not before the provisional.
- Confirm whether non-patent literature is included. It is the main cost driver and where damaging references hide.
- Specify jurisdictions explicitly, matching the markets you actually sell into.
- Ask what is out of scope, which is more informative than what is in it.
- Separate search cost from opinion cost when comparing quotes.
- Share your own search results with the searcher so their hours go further.
- Budget a search at each renewal window too. Against an $8,280 third fee, it is the cheapest evidence available.