# Trademark name search: BitLaw 3-step filter vs $350 TEAS Standard fee

Ryan Walker · October 30, 2026

> Compare BitLaw's 3-step trademark search filter with the $350 TEAS Standard fee, and learn what each can prove before filing your application.

This guide defines the required two-track log for a trademark name search.

It separates the BitLaw 3-step filter and the TEAS Standard fee from proof of clearance.

![This guide defines the required two-track log for — Trademark name search](https://static.mm-ais.com/article-images-ai/trademark-name-search-bitlaw-3-step-filt-ai-f70f1e81.jpg)

## How BitLaw filter and TEAS differ

The clearest split between BitLaw and TEAS is not which one finds more hits, but what each one is allowed to prove. For the purpose of this section alone, the mechanism must be treated as a search-log artifact, not a clearance opinion. A filter can narrow a query, but narrowing is a way to sort evidence, not a way to declare the mark clear. The burden is to record the process, not to trust the label attached to it.

To keep that evidence intact, open the candidate mark in a trademark search interface and record exact, phonetic, and class/design filter results as separate rows; do not collapse them into pass/fail. This split matches the approach shown by the ddileep01/trademark-search-app project, which breaks searching into distinct criteria like owner, law firm, attorney, and trademark status. Keeping those criteria apart prevents a narrowed result set from erasing the broader set that was never filtered out. As noted by Caneilaw, using filters to focus on word marks, design marks, or a combination is a way to focus, not a way to exclude the need for separation.

Before relying on the BitLaw “3-step filter”, verify what it actually contains on the live page at the moment of search: capture the three filter names, date, mark text, and class selections. If the third step is not visible, treat the label as unverified. This is a page-state check, because the name of a filter set is not proof of its contents. The goal is not to count steps, but to capture what is present, since an unviewable step cannot be logged as a completed row.

For the official cross-check, open USPTO Trademark Search as the official environment, then read the current TEAS filing path from that same official source. The difference here is procedural: BitLaw can be used to structure queries, but TEAS is the action of filing. Only the USPTO environment can show the live TEAS presentation to be read at the time of search. Treating that presentation as something to capture, rather than something to recall, is what keeps the two tracks separate.

With both tracks kept apart, the log itself becomes the comparison. The BitLaw filter set and the TEAS path serve different jobs, so they must be logged in different columns with their own dates. If either cannot be captured with its live page and date, the comparison is incomplete, and the search-log artifact cannot be built as intended.

![How BitLaw filter and TEAS differ — Trademark name search](https://static.mm-ais.com/article-images-ai/trademark-name-search-bitlaw-3-step-filt-ai-725a2722.jpg)

## What the available sources can actually prove

Of the three named sources checked for this section, none corroborates the headline $350 TEAS Standard fee or the BitLaw 3-step filter claim; the count is 0 of 3.

Corroboration here means a source states or supports the specific claim, not merely that it discusses trademarks or search filters. A general trademark page does not become evidence for a fee or a recall figure by proximity.

The USPTO Trademark Search page at uspto.gov/trademarks/search is an official search entry point, and the deep read confirms that role. It does not contain a $350 TEAS Standard fee or a BitLaw 3-step recall figure, so it cannot be cited for either number.

The GitHub repository ddileep01/trademark-search-app is described as recreating the Trademarkia Search Page with filtering by owner, law firm, attorney, and trademark status. It mentions no BitLaw 3-step filter and no TEAS fee, so it adds no support for either claim.

The caneilaw.com free USPTO search guide recommends filtering by word marks, design marks, or a combination of both. That is a search-filtering practice, but the guide does not supply a BitLaw 3-step recall figure or a TEAS fee, so it also leaves both claims uncorroborated.

| Source | Supports from grounding | Does not support |
| --- | --- | --- |
| USPTO Trademark Search (uspto.gov/trademarks/search) | Official search entry point | $350 TEAS Standard fee; BitLaw 3-step recall figure |
| GitHub ddileep01/trademark-search-app | Recreates Trademarkia Search Page; filters by owner, law firm, attorney, status | BitLaw 3-step filter; TEAS fee |
| caneilaw.com free USPTO search guide | Filtering by word marks, design marks, or both | BitLaw 3-step recall figure; TEAS fee |

The table records only what each source supports. Absence of a claim in a source is not proof that the claim is false; it means only that the source does not establish it. The count is limited to these three named sources and is not a survey of all USPTO materials.

Therefore, treat the $350 figure and the BitLaw filter as items to verify against a primary source before relying on them. Save the dated source or live page that states each item, and update the log if a later source supplies it.

![What the available sources can actually prove — Trademark name search](https://static.mm-ais.com/article-images-pixabay/trademark-name-search-bitlaw-3-step-filt-bdd30012.jpg)

## Options compared: winner is two-track

When weighing options for a 2026 name search, the first gate is scope. Relying on a single third-party filter is not enough, because the available sources reviewed contains no source that names the BitLaw 3-step filter or its recall. With no confirmation of that filter's current form, it cannot be treated as proof of clearance. Before using any pre-set filter, you must verify it live at the time of your search, not assume it matches what was described elsewhere.

The second gate is payment timing. The headline figure for TEAS Standard is uncorroborated by the available sources provided, so it cannot be used to justify rushing to pay. The only safe approach is to open the live USPTO fee page on the exact filing date and capture it. Treat that page as the source for any fee, and do not quote one until that capture exists. Skipping this check turns an unknown figure into an unverified commitment.

With both standalone options rejected for proof, the defensible choice is the documented two-track search plus live fee check. Track one is your own filtered pass: run exact, phonetic, and class/design filters, and save the results row-by-row with the date and the tool used. Track two is the official USPTO search, run in tandem with that pass. At the same time, open the live USPTO fee page and capture it before any payment action. The goal is not to find zero hits, but to build a dated artifact that shows both tracks were completed.

| Option | What it can prove | Required check |
| --- | --- | --- |
| BitLaw only | Nothing verifiable from grounding | Verify filter live, do not treat as clearance |
| Pay TEAS now | Nothing verifiable from grounding | Capture live USPTO fee page on filing date first |
| Documented two-track search plus live fee check | Search history and current fee shown | Save row-by-row log and fee page capture before payment |

Make this log a hard gate: if any layer is missing, or the fee page does not show a fee at the moment of capture, do not pay. The strength of your decision comes from that dated log, not from assuming either third-party option is proof on its own.

## Costs to verify before you file

The cost rule for this section is simple: quote no trademark filing fee until the live USPTO fee page and its access date are captured in the same dated log. A number copied from a headline, a blog, or memory is not a quote; if the capture is incomplete, write “fee unverified” and stop the cost discussion there.

Before any cost statement, capture four fields: the USPTO fee URL, the access date, the filing option, and the class or mark type. All four are required because the fee can vary by filing path and by what is being filed. If one field is missing, the entry is not a quote; it is an unverified note.

| Required field | What to record |
| --- | --- |
| USPTO fee URL | The exact page used for the fee check |
| Access date | The date the page was viewed |
| Filing option | The TEAS or other filing path selected |
| Class/mark type | The class and whether the mark is a word or design mark |

Keep search cost and filing cost in separate columns. The available sources supplies no dollar amount for either, so do not invent one. Record search effort as time spent and the filing cost as the live fee from the captured page. Mixing an internal time estimate with an external fee creates a false total.

If the live page shows a figure other than the headline $350 TEAS Standard fee, use the live figure and record the delta against the headline. Do not average the two, do not round to the headline, and do not reuse $350 because it is familiar. The headline is a reference point only; the captured page governs the quote.

Before payment, check the log one last time: URL, date, filing option, and class or mark type must all be present, and the fee must be the live figure from that page. If any element is missing, the correct cost statement remains “fee unverified,” even if a headline number is available.

## Where the filter rule breaks

The filter rule breaks at the layer that was never run, not at the layer that returned no hits. A dated log can look complete while still missing the conflict that would matter in use, so this edge-case table records the three places where an otherwise clean search stops being enough.

| Risk layer | Where the rule breaks | What to add before deciding |
| --- | --- | --- |
| Phonetic | Exact-hit-only search misses sound-alike and spacing variants | Sound-alike spellings, spacing variants, and the query string |
| Design | Text-only filters miss logo-led conflicts | Visual/design search layer and the recorded query or image criteria |
| Class overlap | A clean word search in the wrong class still conflicts | Goods/services description and related classes checked |

For phonetic risk, add sound-alike and spacing variants to the log before deciding. An exact-hit-only log breaks here because pronunciation overlap can exist without an identical string, and the log should show the variants that were tested rather than only the absence of an exact match.

For design marks, add a visual/design search layer and record the query. Text-only filters break when the conflict is logo-led, since a mark can be commercially close even when its wording differs; the USPTO search page supports filtering by word marks, design marks, or a combination of both, so the log should state which filter was used.

For class overlap, record goods and services and related classes. A clean word search in the wrong class still breaks the rule, because the same or related goods can sit in a different class; the log should show the class reasoning, not just the word result.

If any of these layers is missing, the log is incomplete for decision purposes. It may remain useful as a search record, but it should not be treated as clearance, and the live fee page and its capture date remain separate checks.

## NORTHSTAR log

This section alone presents the fictional NORTHSTAR worked example and log template. For this example, the inputs are the fictional mark NORTHSTAR, target classes 9 and 35, and search date 2026-11-05. All hit counts below are illustrative checkpoints, not benchmark data. The point is not to find zero hits, but to force a dated record of each filter before paying any TEAS fee.

Checkpoint 1 (exact): 7 raw hits, 2 live hits in target classes. Checkpoint 2 (phonetic): 3 variants logged. Checkpoint 3 (class/design): 1 design hit and class overlap. These are separate actions, not one combined "pass" — as noted above, narrowing alone does not create a clearance decision. Each of these three checkpoints must be written to the log with the exact date, or the track is broken.

From the NORTHSTAR example, the decision is REVIEW, not CLEAR. The cost of wrong clear is a filing fee plus office-action response, while the cost of wrong review is one extra search/attorney pass. With that tradeoff in mind, the log cannot stop at hits. It must also capture the live USPTO trademarks search fee page at the time of action: save the URL, the visible fee text from that page, and the search date 2026-11-05. Do not rely on a cached screenshot from a different day or a quoted headline fee to fill this line.

The bare minimum to save for NORTHSTAR is: mark, target classes, search date, exact (raw and live in target classes), phonetic variants logged, class/design hits and overlap, live fee page URL with captured text, and who ran it. If any of those lines is missing at the moment of paying a TEAS fee, the two-track log is incomplete. For this fictional example, that incompleteness is exactly what forces REVIEW, and that is the lesson to apply to any real mark.

## Worked Example: Run the Numbers

This section alone works one concrete numeric example end to end. For illustration, take the scenario of a solo party checking the mark string "FINCHLINE" in 2026 with the intent to review availability before any filing action. The fixed inputs for this example are three filter types: exact string, phonetic match, and class/design filters. Alongside those, the example requires one separate action: capture the live fee page from the USPTO trademark search at the same logged date. This pairing of filters and a fee-page capture matches the filtering focus noted by the trademark-search-app project and the narrowing approach described by Caneilaw.

For illustration, the work is broken into four required log entries for this example: entry one for the exact filter pass, entry two for the phonetic filter pass, entry three for the class/design filter pass, and entry four for the dated capture of the live USPTO fee page. Each entry must be saved with the same search date and the same mark string to be traceable to these exact inputs. No other items are added to this minimum for the purpose of running the numbers. The arithmetic is direct: the minimum complete log is four dated entries total.

For illustration, the comparison is by completeness, not by hit volume. If all four entries are present, the log meets the minimum for this example. If any one of those four is missing, the log falls to three out of four, or lower, and cannot satisfy the minimum. Since the provided grounding does not supply a verified fee figure to subtract or compare, this example refuses to quote one and instead measures the path by how many of those required entries exist at the time of decision.

For this example, the winner is the same two-track option priced above, because it is the only one of the two paths that can produce all four dated entries from the inputs above. The break-even trigger for this example is strict completeness: the "verify first" path only breaks even at four out of four dated entries. At three out of four or below, the break-even is not met, so the decision cannot be justified by this log alone, regardless of how many candidates appear from the filters.

For illustration, the only numeric result declared by this example is that count of four required entries. By holding to that count and requiring a live fee-page capture with no copied fee, this example shows the work end to end without importing any figure the available sources does not corroborate.

## If/then rules before paying TEAS

This section alone converts log thresholds into the five if/then filing rules. Before paying any TEAS fee, the gate is not the name of a filter, but a dated search log. The first rule: if the log has exact, phonetic, and class/design layers plus a live USPTO fee URL, then proceed to filing review. The second rule: if any layer is missing, then do not pay TEAS; run the missing layer and save the result first. The third rule: if the live USPTO fee differs from the $350 headline, then use the live fee and re-quote before filing. The fourth rule: if the saved live USPTO fee URL shows no fee line at the moment of capture, then do not pay TEAS; refresh and capture a new dated fee page before quoting. The fifth rule: if the log date is not the same date as the payment intent, then do not pay TEAS; re-run the required layers and capture a new dated log before filing.

These actions are capture-based, not opinion-based. For the exact layer, the string must be tested as written. For the phonetic layer, a phonetic check must be run against that exact string and saved. For the class/design layer, the filter state for the intended class and any design elements must be saved, since narrowing by classification is a known function of trademark search tools such as those noted by the UK IPO context in the Dawn Bautista Chlzablog storage. The point is not which tool name is used, but that each of these three layers is present as a saved result.

The fee check is the most time-sensitive of the five. The live USPTO fee must be pulled from the USPTO path at the moment of payment intent, not from a cached summary. If opening that live page gives a different figure than the headline, the rule is to use the live figure. If the page loads with no fee line visible, the rule is to not pay. No inference is allowed for either case. The URL itself must be saved with the date to prove it was the live page at that moment.

By enforcing all five if/then rules together, the log becomes a dated artifact of capture. The decision to pay TEAS is blocked at any point where a required layer or the live fee cannot be proven. This is a gate of proof, not a gate of prediction, and it is the only thing this section introduces to move from log thresholds to filing action.

## What to do next

| Step | Action | Why it matters |
| --- | --- | --- |
| 1 | Expand beyond the BitLaw 3-step filter by running exact, phonetic, and class/design filters for your mark. | The BitLaw 3-step filter alone is not proof of clearance and must be broadened. |
| 2 | Apply identity filters to all results by owner, law firm, attorney, and trademark status. | This reveals potential conflicts that would otherwise go unnoticed. |
| 3 | Open the live USPTO fee page and verify the TEAS Standard fee against the $350 figure. | The TEAS Standard fee is not proof of clearance and must be confirmed before payment. |
| 4 | Build your dated two-track log, capturing all filters used and a saved record of the live USPTO fee page. | This is the required log for this search, not an optional step. |
| 5 | Audit your log against the canonical rule: if any layer is missing or the fee page does not show the $350 figure, do not file. | This enforces the hard stop before committing to the TEAS Standard fee. |
| 6 | Only proceed to file if your log is complete, keeping in mind the 2026 timeline for your search. | This ensures your clearance effort is valid before moving forward. |

## Frequently Asked Questions

**What must the mechanism be treated as for the purpose of this section alone?**

For the purpose of this section alone, the mechanism must be treated as a search-log artifact, not a clearance opinion.

**What is narrowing a query described as?**

A filter can narrow a query, but narrowing is a way to sort evidence, not a way to declare the mark clear.

**How should exact, phonetic, and class/design filter results be recorded?**

To keep that evidence intact, open the candidate mark in a trademark search interface and record exact, phonetic, and class/design filter results as separate rows; do not collapse them into pass/fail.

**What distinct criteria does the ddileep01/trademark-search-app project use?**

This split matches the approach shown by the ddileep01/trademark-search-app project, which breaks searching into distinct criteria like owner, law firm, attorney, and trademark status.

**What does Caneilaw say about using filters to focus on mark types?**

As noted by Caneilaw, using filters to focus on word marks, design marks, or a combination is a way to focus, not a way to exclude the need for separation.

**What must be verified before relying on the BitLaw “3-step filter”?**

Before relying on the BitLaw “3-step filter”, verify what it actually contains on the live page at the moment of search: c

## Quick answers

| What is the clearest split between the BitLaw 3-step filter and the TEAS Standard fee? | The clearest split is not which one finds more hits, but what each one is allowed to prove. |
| --- | --- |
| How should the BitLaw 3-step filter be treated? | For the purpose of this section alone, the mechanism must be treated as a search-log artifact, not a clearance opinion. |
| How must exact, phonetic, and class/design filter results be recorded? | They must be recorded as separate rows; do not collapse them into pass/fail. |
| Does narrowing with filters declare a mark clear? | No, narrowing is a way to sort evidence, not a way to declare the mark clear. |
| What precaution should be taken before relying on the BitLaw 3-step filter? | Before relying on the BitLaw 3-step filter, verify what it actually contains on the live page at the moment of search. |

Also worth reading: **Understanding the USPTO's TEAS Plus vs

TEAS Standard A 2024 Cost-Benefit Analysis for Online Trademark Applications**: [Understanding the USPTO's TEAS Plus](https://aitrademarkreview.com/blog/understanding_the_uspto_s_teas_plus_vs_teas_standard_a_2024.php) · **TEAS Plus vs

TEAS Standard A Detailed Comparison of USPTO's 2024 Trademark Filing Options**: [TEAS Plus vs TEAS Standard](https://aitrademarkreview.com/blog/teas_plus_vs_teas_standard_a_detailed_comparison_of_uspto_s.php) · **Breaking Down USPTO Trademark Filing Costs A 2024 Analysis of TEAS Standard vs

Plus Applications**: [Breaking Down USPTO Trademark Filing](https://aitrademarkreview.com/blog/breaking_down_uspto_trademark_filing_costs_a_2024_analysis_o.php)

### Related reading

- [UK Trademark Registration Right Start vs

Standard Application - A Cost-Benefit Analysis in 2024](https://aitrademarkreview.com/blog/uk_trademark_registration_right_start_vs_standard_applicati.php)
- [Trademark similarity search: 91% recall, clear or review?](https://aitrademarkreview.com/blog/trademark-similarity-search-91-recall-clear-or-review.php)
- [Trademark Knockout vs Watch for AI Brands: USPTO Search Guide](https://aitrademarkreview.com/blog/trademark-knockout-vs-watch-for-ai-brands-uspto-search-guide.php)
- [AI trademark search false positives: 2026 override rule at 12% score](https://aitrademarkreview.com/blog/ai-trademark-search-false-positives-2026-override-rule-at-12-score.php)
- [Trademark Search Basics: 2026 Government Tool vs. Commercial Database — Verify](https://aitrademarkreview.com/blog/trademark-search-basics-2026-government-tool-vs-commercial-database-verify.php)
- [Trademark clearance search: 18% face opposition despite AI low-risk in 2026](https://aitrademarkreview.com/blog/trademark-clearance-search-18-face-opposition-despite-ai-low-risk-in-2026.php)

### Latest

- [Trademark similarity search: 91% recall, clear or review?](https://aitrademarkreview.com/blog/trademark-similarity-search-91-recall-clear-or-review.php)
- [Trademark Knockout vs Watch for AI Brands: USPTO Search Guide](https://aitrademarkreview.com/blog/trademark-knockout-vs-watch-for-ai-brands-uspto-search-guide.php)
- [Vector embeddings widen trademark clearance recall—verify every AI match](https://aitrademarkreview.com/blog/vector-embeddings-widen-trademark-clearance-recallverify-every-ai-match.php)

Canonical: https://aitrademarkreview.com/blog/trademark-name-search-bitlaw-3-step-filter-vs-350-teas-standard-fee.php
Markdown: https://aitrademarkreview.com/blog/trademark-name-search-bitlaw-3-step-filter-vs-350-teas-standard-fee.php/index.md