TL;DR
COMLEX-USA Level 1 is not Step 1 with an osteopathic sticker on it. The NBOME blueprint reserves a minimum of 12% of items for Osteopathic Principles, Practice, and Manipulative Treatment and another 6% for osteopathic patient care, weights the musculoskeletal system above every other clinical presentation, and writes vignettes that read differently from NBME items. The tools DO students already use are good at what they were built for: COMBANK and COMQUEST are the COMLEX-native question banks, UWorld and AMBOSS cover the shared biomedical core with modest OMT supplements, Savarese and your school’s OMM notes carry the manipulative-medicine content, and the NBOME’s own COMSAE is the only self-assessment written by the people who write the exam. Where they stop is where every fixed resource stops: they cannot write questions on the technique your school emphasized, turn your OMM lab notes into a spaced-repetition deck, or answer a 2 a.m. question about Chapman’s points with a citation. That is the job of an AI layer, and this guide shows where to add one without pretending it replaces a question bank. To be clear up front: Neural Consult’s perfect score on published USMLE sample sets is a USMLE benchmark, not a COMLEX result, and we do not claim otherwise.
Why Level 1 needs a different plan than Step 1
If you are an OMS2 in 2026, you are preparing for an exam that changed shape this year. The NBOME announced in December 2024 that Level 1 would drop from 352 items to 320 starting with the May 2026 testing cycle, and its current description of examinations now reads accordingly: 320 single-best-answer questions in eight sections of 40, split across two separately timed four-hour sessions, with 60 minutes of pooled break time you can spend after any section. Seat time stayed at eight hours, so each question gets a little more room. Level 1 has been reported as pass or fail since 2022, and passers receive a Formative Performance Profile comparing their content-area performance with other first-time passers. With no three-digit score to optimize, the goal becomes “leave no blueprint category exposed.”
That blueprint is why a Step 1 plan is not enough. The Level 1 blueprint has two dimensions. Dimension 1 is competency domains: Osteopathic Principles, Practice, and Manipulative Treatment carries a minimum of 12% of items, Osteopathic Patient Care and Procedural Skills another 6%, and Application of Knowledge for Osteopathic Medical Practice 60%, three quarters of which is the foundational biomedical science you share with MD classmates. Dimension 2 is clinical presentations, where the musculoskeletal system leads at a minimum of 13%, ahead of community health and wellness at 12% and the nervous, gastrointestinal, circulatory, and respiratory systems at 10% each. Those are floors, not targets, so OPP and MSK together can run well above a fifth of your exam. A resource built for Step 1 has no reason to cover Fryette’s principles, sacral diagnosis, counterstrain tender points, or viscerosomatic reflexes, and most do not.
The second reality is that many DO students sit both exams. Step 1 is also pass or fail, and the biomedical core overlaps heavily, which is why “study for Step 1, add OMM” is such a common plan. It mostly works. The failure mode is treating OMM as a two-day bolt-on at the end of dedicated when the blueprint makes it the largest competency domain outside general knowledge application. National first-time pass rates on Level 1 have run at approximately 90% in recent cycles, according to the outcome tables osteopathic schools publish from NBOME data. Put the other way, roughly one in ten first-time takers does not pass, and those students are rarely the ones who never opened a question bank. More often their preparation had a gap the blueprint was always going to find.
The tools DO students already use, and where each one stops
None of this is an argument to drop a resource you are using well. It is a map of what each one was built for, so you see the gaps before the exam does.
COMBANK (TrueLearn)
COMBANK is the COMLEX-native question bank most DO schools point students toward. TrueLearn lists more than 2,900 NBOME-formatted items mapped to the Level 1 blueprint, hundreds of OMM videos, a Prometric-style interface, and national benchmarking, with the SmartBank priced from $199 for 30 days to $549 for a year and a 160-question predictive assessment as an add-on. Its real value is that the questions are written in the NBOME’s voice, which differs noticeably from the NBME’s, and that OMM is treated as a core subject rather than an appendix. Two cautions. TrueLearn’s marketing claims, including that more than 93% of OMS2 students use COMBANK and that students improve by an average of 60 points, are the company’s own figures without a published methodology, and the second refers to a three-digit score Level 1 no longer reports. And like every fixed bank, COMBANK cannot write a new question on the exact variant of a technique your OMM department tested last block.
COMQUEST
COMQUEST is the other COMLEX-specific bank, with more than 2,040 Level 1 questions and explanations covering high-yield OMT as well as the medical jurisprudence and ethics items that catch Step-trained students off guard. Students who use both banks often describe COMQUEST as closer in difficulty and tone to the real exam, though that is reputation rather than data. COMQUEST does not publish pricing on its product page, so we will not quote a figure. Its limitation is COMBANK’s: a finite pool of items written to the national blueprint, not to your curriculum.
UWorld with the OMT supplement
If you are taking both exams, UWorld’s Step 1 bank is probably already in your plan, and its explanations remain the standard the market is measured against. For Level 1, UWorld bundles its 3,600-plus Step 1 questions with a COMLEX-style OMT supplement of 140-plus questions, priced from $369 for 30 days to $599 for 360 days, or sells the OMT supplement alone for $49 to $99. Be clear-eyed about what 140 questions can do: it is a calibration set for the OMM item style, not a comprehensive OMM review, and UWorld’s page describes the content only as “COMLEX-style” without listing what it covers. We have written at length about where UWorld wins and where an AI layer complements it; for DO students, the short version is that UWorld covers the 75% you share with MD students extremely well and leaves most of the osteopathic remainder to someone else.
AMBOSS
AMBOSS is a library first and a question bank second. For COMLEX it lists osteopathic modules within its 1,300-plus articles, a high-yield OMM study plan, and 100-plus COMLEX-style OMM questions inside a 5,700-plus question bank spanning all three COMLEX levels, at $378 for six months or $448 for twelve, with the unlimited bank sold as a separate upgrade. It is a useful reference to read during question review and a reasonable place to check a principle you half-remember. It is not where most DO students get OMM question volume; 100-plus items is a supplement, not a bank.
Savarese, your school’s OMM notes, and Anki
Ask any OMS3 what they used for OMM and the answer usually includes the green book: Savarese’s OMT Review, the compact outline passed down through osteopathic classes for years, dense, organized by region and technique, and ending in practice questions. Pair it with your school’s OMM slides and lab handouts, because the blueprint is national but your faculty’s emphasis is the best predictor of what you will remember in the room. Most students then move that content into Anki, from a shared OMM deck or by making cards. The problem is the hours. Anki is the best scheduler available, free, and worth keeping; building a thorough OMM deck from scratch is a weekend you may not have during dedicated.
COMSAE Phase 1
The NBOME’s own COMSAE Phase 1 is built to the same specifications as Level 1 and is the only practice exam written by the organization that writes the real one. Take it. Its limit is that it is a snapshot, not a study tool, so use it the way you would an NBME self-assessment for Step 1: to find the blueprint categories where you are exposed, then fix them somewhere else.
Where an AI layer earns its place for Level 1
Every tool above shares one property: its content was finished before you opened it. That is fine for the biomedical core, where the banks are deep and the material is stable. It is a real limitation for the osteopathic fifth of the exam, where the banks are shallower, the item style more idiosyncratic, and the best source of truth is your own school’s teaching. An AI layer earns its place by doing four specific jobs fixed resources cannot, and should be judged on those jobs rather than on vague promises about personalization.
Job 1: Unlimited practice on the OMM topics the banks run thin on
When you have done every counterstrain question in COMBANK and still feel shaky on lower-extremity tender points, a fixed bank has nothing left to give you. Neural Consult’s Question Generator writes new board-style items on any topic at any difficulty, with explanations, so you can ask for twenty vignettes on sacral torsion diagnosis, viscerosomatic reflex levels, or rib dysfunction and get a fresh set each time. Two honest caveats. First, it is a general medical question generator you steer by topic, not an NBOME item pool, so hold its OMM items against Savarese and your school’s notes as you would any third-party source. Second, our accuracy evidence is a perfect score on the published USMLE Step 1, Step 2 CK, and Step 3 sample sets, plus a blinded study in which physician raters scored our generated questions equal or superior to NBME items, including 4.4 versus 3.5 on explanation quality (p<0.001). That is strong evidence the generator understands the shared biomedical core. It is a medicine benchmark, not a COMLEX benchmark, and nobody has run the equivalent study against NBOME items.
Job 2: Turning your OMM lectures into a deck you did not have to build
Your school’s OMM slides are the highest-yield osteopathic source you own and the one no commercial product has seen. Upload them to the AI Lecture Notebook to ask questions of your own course material, then push the content into the Flashcard Hub, which builds spaced-repetition decks automatically and exports to Anki if that is where you already live. The point is not that AI-made cards are better than hand-made ones. The best independent evidence on that question, a medRxiv preprint from Brown’s Warren Alpert Medical School that tested a similar tool (not Neural Consult) with first-year students, found that AI-generated cards and summaries saved time, with 74% of students reporting savings from flashcards and 61% from summaries, but produced no significant difference in exam scores. Read that honestly: the gain is hours, not points, and the hours only matter if you spend them on questions.
Job 3: Cited answers instead of a confident guess
The temptation during dedicated is to ask a general chatbot the thing you cannot find in the book. For OMM that is risky in a specific way: the osteopathic literature is smaller than the general medical literature, which gives a general-purpose model less to draw on and more room to improvise. A 2024 study in JMIR found that roughly 28.6% of the references one widely used chatbot produced were fabricated, and a 2025 paper in Communications Medicine found that leading models repeated planted clinical errors up to 83% of the time. Neural Consult’s AI Medical Search uses a grounded-retrieval architecture that answers with citations, including PubMed, so when you ask which spinal levels carry sympathetic innervation to the appendix, you get a source you can open rather than a sentence you have to trust. We have a longer piece on what the evidence does and does not show about AI study tools if you want the full picture first.
Job 4: Covering the dual-exam overlap without paying for it twice
If you are sitting Step 1 and Level 1 within weeks of each other, the biomedical core is the same material under two question styles. A generator you steer by topic lets you practice renal physiology or pharmacology in vignette form without buying a second bank, and ask for an osteopathic framing when you want to switch gears. It does not replace the COMLEX-native banks for final calibration to the NBOME’s item style; use COMBANK or COMQUEST for that in the last weeks. And one feature we will not oversell: the Case Simulator, which runs voice-enabled, rubric-scored patient encounters, is a clinical-years tool for history-taking and reasoning, useful for Level 2 and OSCEs. It is not a Level 1 tool, and we would rather say so than pad the list.
A dedicated-period shape for Level 1
Dedicated periods vary by school, with four to eight weeks common. This template assumes six, scales either way, and borrows its skeleton from our Step 1 dedicated schedule guide, with the osteopathic content deliberately moved from the end to the beginning.
Weeks 1 and 2: Build the OMM floor first
Read Savarese once, front to back, alongside your school’s notes. Upload the OMM lectures to the Lecture Notebook and generate your deck so the review cycle starts on day one rather than week five. Begin COMBANK or COMQUEST in tutor mode by system, mixing an OMM block into every day, and use the Question Generator for the technique families you keep missing, checking its explanations against the book. Sixty to eighty questions a day is a reasonable start for most students; the number matters less than finishing every block with a written note on why you missed what you missed.
Weeks 3 and 4: Shared core, osteopathic framing
This is where the Step 1 overlap pays off. Move through the biomedical systems in whichever bank you trust most, keep the daily OMM deck reviews going, and sit your first COMSAE Phase 1 at the end of week 4 to see which blueprint categories are exposed. Spend the next two days on the weakest two, with generated questions aimed at exactly those topics.
Weeks 5 and 6: Calibrate to the NBOME’s voice
Shift to timed, random, mixed blocks in a COMLEX-native bank, 40 questions at a time to match the new section length, and use the generator only for targeted gap-filling, not volume. Take a second COMSAE in the final week if you can, taper your question count in the last three days, and spend the final evening on the OMM deck rather than anything new. If both exams fall close together, schedule Level 1 second where possible, because the osteopathic material decays fastest and benefits most from being freshest.
The recommended stack
| Job | Best tool | Why | Where Neural Consult fits |
|---|---|---|---|
| COMLEX-native question volume | COMBANK or COMQUEST | Written in the NBOME’s voice, mapped to the blueprint, with OMM treated as a core subject | Fills topic gaps the banks run thin on with unlimited generated items |
| Shared biomedical core (dual exam) | UWorld, plus its OMT supplement if you want style calibration | The deepest explanations on the 75% you share with Step 1 | Topic-steered practice without a second bank; osteopathic framing on request |
| OMM content review | Savarese and your school’s OMM notes | Dense, region-by-region, and your faculty’s emphasis is the best predictor of what you will recall | Lecture Notebook and Flashcard Hub turn those notes into a deck in minutes |
| Reference during review | AMBOSS or First Aid | Readable articles you can check mid-block, including osteopathic modules | AI Medical Search for cited answers when the article does not cover it |
| Spaced repetition | Anki | Free, proven scheduler; keep it if you already use it | Flashcard Hub exports to Anki so you do not rebuild your workflow |
| Self-assessment | COMSAE Phase 1 | The only form written by the people who write the exam | Generated questions aimed at the categories COMSAE exposes |
| Clinical-years simulation | Not a Level 1 need | Save it for Level 2 and OSCEs | Case Simulator, later |
The bottom line
The students who struggle with Level 1 are rarely the ones who skipped the question bank. They are the ones who prepared for Step 1 and assumed the osteopathic fifth of the exam would take care of itself. Keep COMBANK or COMQUEST at the center of your plan, Savarese and your school’s notes as the source of truth for OMM, and Anki if it is already working for you. Then add an AI layer for the jobs fixed resources cannot do: practice on the techniques the banks run thin on, a deck built from your own OMM lectures, and a cited answer instead of a confident guess when the book runs out. Neural Consult is used by more than 75,000 health professions students and was named Best AI-Powered Medical Learning Platform 2026 – USA by the GHP Healthcare & Pharmaceutical Awards; neither fact is a COMLEX pass rate, and we will not pretend otherwise. The free tier is enough to test whether generated OMM questions and a lecture-built deck earn a place in your plan before you spend anything.
Try Neural Consult free and see whether an AI layer fits your Level 1 plan →