The Astrophysicist who proves the simulation is right
$296,740top of the range in California · middle $172,250 / yr
AI augments this role
Astrophysicists in the United States earn a median of $172,250 a year. Pay starts near $82,110. Pay reaches $296,740 at the top of the range in California, the best-paying state for this work among those with at least 500 people in the job.
Source: U.S. Bureau of Labor Statistics, Occupational Employment and Wage Statistics, May 2025 (Physicists, SOC 19-2012). Last checked 9 September 2026.
Entry level
$82,110
Top of the range · California
$296,740
Education
Doctoral degree in Astrophysics
Wages — U.S. Bureau of Labor Statistics, Occupational Employment and Wage Statistics, May 2025 (Physicists). Top of the range is the highest state-level figure among states with at least 500 people in the job. AI-impact rating is PayCrunch's editorial assessment. Updated September 2026.
🆕 New & Trending AI Tools for AstrophysicistReviewed September 2026
We track new AI-tool launches every week and refresh this list — here’s what’s gaining traction for Astrophysicist work right now.
Julius AINEWFree / $20 mo
AI data analyst that runs statistics and charts from plain-language prompts.
How an Astrophysicist uses it: analyze datasets and generate figures without writing code
NotebookLMNEWFree / $7.99 mo
Google tool that answers questions grounded only in the documents you give it — with citations.
How an Astrophysicist uses it: load your own manuals, policies, or PDFs and ask questions that stay accurate to the source
ElicitFree / $12 mo
AI research assistant that finds and summarizes papers.
How an Astrophysicist uses it: run a literature review and extract findings across dozens of papers fast
ConsensusFree / $9 mo
AI search that answers questions from peer-reviewed research.
How an Astrophysicist uses it: get evidence-backed answers with the studies behind them
SciSpaceFree / paid
AI that explains papers and helps with literature review.
How an Astrophysicist uses it: decode dense papers and trace citations quickly
SciteFree / $20 mo
Shows whether other studies support or contradict a paper's claims (Smart Citations).
How an Astrophysicist uses it: check if a finding is actually backed by the wider literature before you cite it
ChatGPTFree / $20 mo
The most-used AI assistant — writing, analysis, research, and images from a plain-language chat.
How an Astrophysicist uses it: draft emails and documents, summarize long files, and get instant answers to on-the-job questions
ClaudeFree / $20 mo
AI assistant known for careful writing, long-document analysis, and coding.
How an Astrophysicist uses it: analyze big reports or spreadsheets and turn messy notes into clean, finished writing
Google GeminiFree / $20 mo
Google's AI assistant, built into Gmail, Docs, and Search.
How an Astrophysicist uses it: draft and reply inside Google Workspace and research without leaving the page
A simulation has been running since dawn and the curve on the screen still misses the catalog. The astrophysicist changes one assumption about how matter and radiation interact, reruns the slice that failed, and writes down why the previous version was too kind to the data. This is theory and modeling pointed at the universe: equations, code, and a hard comparison to measurements. The night on a telescope belongs to a neighboring kind of work. The seat described here is the physics.
Models that have to survive the sky
An astrophysicist tries to say how the universe behaves. Gravity, light, hot gas, dense stars, galaxies as systems, and the large-scale structure of space are typical problems. Some of the work is pencil-and-paper theory. More of it, in current groups, is numerical. A code evolves a system forward. A statistical test asks whether the fake sky and the real sky can be told apart. The output is a claim with a stated uncertainty, written so another physicist can rebuild the argument.
The day is desks, clusters, and blackboards. Morning might be a derivation that will not close, or a job that died overnight because a boundary condition was sloppy. Afternoon might be a group meeting where a student defends a plot and a senior person asks what happens if a key parameter moves. Observational catalogs enter as inputs. The astrophysicist has to know how those catalogs were made well enough to avoid leaning on a feature that is really the instrument. They do not, in the ordinary week, choose which telescope gets the night. They choose which physical effect is allowed inside the model, and they live with that choice when the paper is refereed.
Collaboration with people who observe is constant and specific. A model that predicts a spectral feature is useful only if someone can measure it. The astrophysicist writes the prediction cleanly enough that an observing team can aim at it, then helps interpret what came back. Authorship, shared code, and who may talk about an unpublished run are settled early. Large collaborations in cosmology and in high-energy work have formal rules. A small theory group has informal ones that are just as real. Either way, the contribution this role is hired for is the physical account, the calculation, and the interpretation, more than the operation of a dome.
Tools are mathematical and computational. Differential equations, statistical inference, simulation codes, and the patience to test a result against a simpler case you can solve by hand. Version control for the code matters because a figure in a paper has to be reproducible a year later when a referee doubts it. Reading is half the job. The literature moves, and a claim that ignores a recent constraint will be dismissed in one paragraph. Writing is the other half. A beautiful model that never becomes a paper does not change what the field believes.
Employers shape the week. A university group mixes research with teaching and with students who are learning to break a problem into pieces. A national laboratory group may tie the science to a facility, a survey collaboration, or a computational program with milestones. A few industry labs hire physicists to model complex systems that are not the sky at all. An astrophysicist who wants to stay on the universe should read the actual project list before falling for a prestigious name. The methods travel. The subject does not always travel with them.
Why these wages say physicist
Same family, this series
The pay figures follow Physicists, SOC 19-2012. That series sits in the same family of physical scientists as astronomers, who are tracked separately. The dollars on this page are the physicist dollars, covering research seats across physics, astrophysics included.
Honesty about the chart changes how you negotiate. A condensed-matter physicist, a particle physicist, and an astrophysicist can all land in SOC 19-2012. The median you see is a median for that whole group. It is fair to use it as the published scale for a research physicist. It is shaky to treat it as a private scale that only sky-modelers earn. When an employer is a physics department, the comparison is natural. When an employer is an observatory that thinks of its staff as astronomers, this chart is still the one this page gives you, and you should say so rather than pretend the Bureau printed a line labeled astrophysicist.
The day-to-day split from observatory observing jobs follows the same honesty. If your week is proposals for telescope nights, live runs, and reduction of your own frames, you are describing a different rhythm from the modeling seat. If your week is equations, simulations, and physical interpretation of data, you are in the work this narrative is about, and the physicist series is the wage series attached to the page. Many careers blend both. Describe the blend in the interview with fractions of time, so the title on the offer and the title on this page are not forced into a false match.
Doctoral preparation, and nothing to hang on the wall
Research seats expect a doctorate. A physics undergraduate degree, with research experience if a lab will have you, is the common start. Doctoral work is a multi-year project under an advisor: courses that make you fluent, then a thesis that adds something the field can use. Astrophysics theses range from analytic theory to large codes to statistical work on surveys. The advisor's taste becomes your early taste. Choose the person for the problems they set and for whether their students finish and find next appointments, not only for the fame of the topic.
There is no licence for this occupation. A department will not ask a state board for permission to hire you. The proof is the thesis, the papers, the code if the work is computational, and letters from scientists who saw the reasoning up close. A research statement that a committee member from another corner of physics can follow will travel farther than a statement written only for your subfield friends. Name the result, the method, and what you want to do next with this employer's students or this employer's machines.
Skills that make you hirable inside physics also make you hirable outside it. Serious computing, careful statistics, and the habit of writing a claim that matches the evidence are valued in industry research and in data-heavy companies. Leaving academic astrophysics is a legitimate decision, especially when permanent research posts are scarce relative to the number of doctorates. Make that decision with open eyes during the doctoral years by talking with people who left, not only with people who stayed. Committees sort candidates by the publication record and the recommendation letters. Those are the constraints that decide the next seat.
Getting a group to take you on
Hiring happens through doctoral admissions first, then through post-degree research openings, laboratory scientist posts, and faculty searches. A doctoral application is transcripts, test requirements the department still lists, a statement, and letters. After the degree, a research opening wants a CV, a statement aimed at that group's science, and letters that mention specific papers. Faculty searches add a job talk and, often, a teaching sample. The talk should show one result in enough detail that a specialist trusts it, then lift the camera so the rest of the department sees why the result matters.
Read the group before you apply. A theory group wants a calculation they cannot already do. A computational group wants code discipline and a scientific target, not only fluency in a language. A group tied to a survey wants someone who can build physical meaning from a catalog. Sending the same statement, with the employer's name swapped, is obvious to readers who see hundreds of files. Mention a paper of theirs you actually understood, and say what you would add. Ask how the salary is funded, whether students are included, and what fraction of the year is yours to define. Those answers predict the job better than the view from the campus.
Interviews probe taste and honesty. You will be walked through a disagreement in the literature and asked where you land. You will be asked what would falsify your favorite model. A candidate who treats every objection as an attack looks risky to live with. A candidate who can change their mind on a chalkboard looks like a physicist. Practical constraints belong in the conversation too: a partner's career, a visa, a need for computing access that the group must provide. Surprises after an offer waste everyone's cycle.
From the thesis to a lasting research seat
The route most research careers follow is graduate student, then a research appointment after the doctorate, then either a scientist post at a laboratory or university or a faculty job. The middle step is where you show the thesis can be followed by a second and third result, often in a new group, often on a clock of a few years. Use it to become the person who frames the problem, not only the person who runs the advisor's code. A second such appointment is common. A long string of them, with no change in independence, is a warning to look at laboratory staff roles, teaching-focused faculty posts, or a move that uses the methods outside astronomy and astrophysics.
Scientist posts at laboratories can be the stable version of the modeling life: a group, a mission or a survey, and a salary that does not depend on winning a new grant every cycle. Faculty posts add courses, committees, and the long responsibility for students whose careers will reflect your supervision. Leadership of a collaboration or a center comes later, and only for people whose science and whose dealings with coauthors both hold up. None of these is a consolation. Each one hires a different fraction of your week for research. Pick with that fraction in view, and keep the physicist identity clear when the employer uses a looser title.
Geography clusters. California, Texas, New York, Maryland, and New Mexico all show up among the stronger median wages on this page, which matches where large physics employers sit, from campuses to laboratories. A move toward one of those states can change the offer more than a repeated year in a thin market. It can also change the science you are near. Go where the group is good, then check the wage against the state median rather than against a story you heard about a coastal campus.
Setting an offer against the physicist scale
The Bureau of Labor Statistics wage release for May 2025 is the source, under Physicists, SOC 19-2012. Entry on this page is $82,110. The median is $172,250. California carries the high figure published for this series, $296,740. The distance from entry to the median is $90,140. The distance from the median to that California figure is $124,490. State medians charted alongside are California $207,490, Texas $184,040, New York $181,970, Maryland $174,390, and New Mexico $165,340. California's median stands $35,240 above the national median.
That $90,140 climb from $82,110 to $172,250 is the central fact for anyone early in the career. A first research appointment after the doctorate can sit near the entry figure even when the work is excellent, because the title is still temporary and the market is crowded. An established physicist, faculty or senior scientist, is the person for whom $172,250 is a meaningful national middle. When you negotiate the first appointment, ask what would move a later offer toward the median, and get the funding term in writing. When you negotiate the established job, the median is the reference, and a gap of tens of thousands below it needs a reason you are willing to accept.
Keep the California numbers sorted. The state median of $207,490 is typical pay for physicists in California on this chart. The $296,740 figure is the high mark published for California, a different kind of number, relevant for senior roles in that state. Using $296,740 as the expected salary for a new research hire will end the conversation. Using $207,490 as a comparison for a California offer, against a national median of $172,250, is the cleaner move. The $35,240 difference between those two middles is large enough to discuss when two employers, one in California and one elsewhere, want the same person. Texas at $184,040, New York at $181,970, and Maryland at $174,390 are all at or above the national median and are fair local anchors. New Mexico at $165,340 sits just under the national median, still in the same conversation.
Startup funds, computing access, student support, and summer salary can change a faculty offer without changing the base. This page reports wages, so price those extras in the employer's own words and keep the base next to $82,110, $172,250, or the state median that fits. A laboratory salary near the California median with stable funding can beat a higher advertised academic rate that depends on grants you have not won yet. Say that plainly. Then return to the model you were hired to build. The offer should fund the physicist. The work should still be about how the universe behaves.
The top of Astrophysicist pay — and how to get there with AI
$296,740what Astrophysicist pay reaches in California
Highest state-level top-of-range annual wage for Physicists, among states with at least 500 people in the job. U.S. Bureau of Labor Statistics, Occupational Employment and Wage Statistics, May 2025.
$82,110entry$172,250middle$296,740top end
What separates the middle of this range from the top of the range is rarely a cleverer model; it is being the one person who can state how wrong the simulation is and defend that figure to a referee.
Designing computer simulations to model physical data is the part of this job that gets rewarded, and checking them is the part postponed until somebody complains. So convergence studies go unrun, resolution effects stay unquantified, and the complex calculations behind a result live in one notebook. Whoever builds the verification layer, meaning regression runs kept in Git, a fixed set of analytic cases, and a written error budget, becomes difficult to replace on any collaboration. Assistants like Claude and GitHub Copilot make that layer cheap enough to build alongside the physics rather than instead of it.
Your playbook, by where you are now
Just startingPin down one result
Take a published simulation result from your group and reproduce it from the input deck alone, logging every undocumented step you had to guess.
Store code and input decks in Git with a tag naming the run behind each published figure.
Add three analytic test cases with known answers and make the code run them on every change.
Do the resolution study nobody did, and express the disagreement in mathematical terms rather than as an aside.
What proves it: A tagged repository where every figure in a paper maps to a run anyone can repeat.
Realistic span: the first couple of years
A few years inTurn checks into infrastructure
Move long runs onto Amazon Web Services AWS software or the local cluster and script the environment with Ansible software so it forms part of the record.
Build a nightly regression comparing fresh output against stored reference runs and reporting the drift.
Write an error budget for one flagship result: discretisation, sampling, input physics, each with a figure and a method.
Have Claude turn your run logs into a draft methods section, then rewrite any sentence the logs do not support.
Present the verification work itself at a scientific conference, because it is a result and almost nobody submits one.
What proves it: An error budget other groups quote when they quote your simulation.
Realistic span: roughly the third through sixth year
ExperiencedSet what counts as verified
Write the verification standard for a collaboration: what a run must ship with before its numbers enter a paper.
Put verification into your research proposals as funded work with named staff attached.
Referee other groups' codes and publish what you find so physics students learn to check their own.
Take the calls where two codes disagree, and settle them with evidence instead of seniority.
What proves it: A published verification standard some collaboration adopted.
Realistic span: from about year seven
The next 90 days
Over ninety days, take the single number from your work you would least like to be questioned on and build its error budget. List each source of error in the calculation: grid resolution, time step, the physical approximations inside the model, the sampling in your analysis of the data. For each, run the cheapest experiment that bounds it, halving a step, doubling a resolution, swapping an approximation, and record what changed. Where you cannot bound it, say so and name the run that would. Write it as two pages, in mathematical terms, with the runs attached, and send it to the colleague most likely to argue. Either you learn the number is soft, which beats a referee learning it, or you hold the one document on the project nobody else could have produced.
Wage figures: BLS OEWS, May 2025. The playbook is PayCrunch editorial guidance, not a guarantee of pay or placement.
Every figure is the national median from the U.S. Bureau of Labor Statistics (OEWS) shown on that role’s own page.
Never used AI before? Start here (2 minutes).
Start with literature and code — the two places AI saves an astrophysicist the most hours. Open NASA ADS for authoritative search and citation graphs, and pair it with Undermind or Elicit to run a deep, cited survey of a subfield in an afternoon instead of a week. For anything you will publish, verify every reference in ADS yourself.
For computation, open GitHub Copilot or Cursor inside your Python or Julia workflow to write and debug analysis and plotting code against the Astropy ecosystem, and use Claude or ChatGPT to explain an unfamiliar method or derivation step by step. Keep the physics judgment yours; let AI remove the boilerplate.
The one rule, forever: Science integrity first. Large language models hallucinate citations, derivations, and numbers — verify every reference in NASA ADS or arXiv, re-derive any equation yourself, and never let AI-generated code touch published results without unit tests and a check against known limits. Disclose AI assistance per journal and funder policy, respect collaboration embargo and data-sharing rules, and never upload proprietary or pre-embargo survey data to a consumer tool.
The plays — exact steps, exact prompts
Do these in order. Each one is copy-paste ready. You do not need to know anything about AI going in.
1
Compress the literature review to an afternoon
Why this pays: Research output is the currency of promotion, grants, and tenure. Finding the right prior work fast — and never missing the paper that scoops or supports you — lets you publish more and write sharper proposals, the direct drivers of a salary at the top of the academic range.
NASA ADSUndermindElicit
1
Use NASA ADS for authoritative search, citation graphs, and export — treat it as ground truth for anything you cite.
2
Run a deep, cited survey of a subfield with Undermind or Elicit.
Copy-paste this prompt
Survey the literature on [e.g., AGN feedback in cosmological simulations and its effect on galaxy quenching] over the last 8 years. Give me the key findings, the main disagreements, the seminal papers, the leading groups, and the open questions — each with citations I can verify. Flag anything that would change how I'd design a new study.
Verify every returned citation in NASA ADS or arXiv — these tools still fabricate references and misattribute findings.
What you'll haveA comprehensive, cited map of your subfield in an afternoon — more papers, sharper proposals, and no missed references.
2
Let AI write the analysis and simulation code
Why this pays: Astrophysics is now computational — the bottleneck is often writing, debugging, and optimizing pipelines, not the physics itself. AI coding assistants cut that time sharply, letting you run more analyses per year and take on more ambitious, fundable projects.
GitHub CopilotCursorClaude Code
1
Use GitHub Copilot or Cursor in your editor to write Astropy/NumPy analysis, vectorize slow loops, and generate publication-quality Matplotlib figures.
2
Debug and accelerate a slow pipeline.
Copy-paste this prompt
Here is my Python analysis code [paste code — no proprietary or embargoed data]. It processes [describe data] and is too slow / returns [wrong result]. Find the bug, then rewrite the hot path to vectorize with NumPy or move it to JAX for GPU. Explain each change and add unit tests against these known limiting cases: [cases].
Never trust AI-written science code blind — add unit tests, check conservation laws and known limits, and confirm reproducibility before publishing.
What you'll haveAnalysis pipelines built and debugged in a fraction of the time — more results per year and room for more ambitious projects.
3
Build ML surrogates and classifiers for your data
Why this pays: Modern astrophysics drowns in data and expensive simulations. Machine-learning emulators and classifiers make inference feasible that wasn't before — the kind of methods paper that gets cited, funded, and noticed for senior roles.
JAXPyTorchscikit-learn
1
Train ML emulators to replace expensive simulation grids (a neural surrogate for a Boltzmann or cosmology code) or classifiers for photometric redshifts and object types, using JAX or PyTorch.
2
Design the ML approach with AI as your methods sounding board.
Copy-paste this prompt
I want to build a neural network emulator for [expensive simulation or observable]. Inputs are [parameters], outputs are [quantity]. Recommend an architecture, a training strategy, how to quantify uncertainty so it's usable for scientific inference, and the validation tests a referee will demand. Note the failure modes I should watch for.
An emulator is only publishable with honest uncertainty quantification and out-of-distribution checks — validate against held-out full simulations, never against itself.
What you'll haveInference and methods that weren't feasible before — the cited, fundable work that defines a senior scientist.
4
Write fundable proposals and clearer papers
Why this pays: Grants and telescope-time awards fund the research and the salary line; clearer writing wins both and moves papers through review faster. AI is a tireless editor and devil's-advocate reviewer for every proposal and manuscript.
ClaudeChatGPTOverleaf
1
Draft in Overleaf, then use Claude or ChatGPT to tighten the science case, write plain-language broader-impacts summaries, and stress-test the logic.
2
Pre-review your own proposal before the panel does.
Copy-paste this prompt
Act as a skeptical NSF/NASA review panelist in [subfield]. Here is my proposal's science justification and methods [paste text — no confidential collaborator data]. List the weaknesses a panel would raise, the missing controls or feasibility concerns, and the 5 concrete edits that would most improve the score. Be harsh.
Use AI to sharpen your own argument — never fabricate preliminary results or citations, and follow the funder's AI-use disclosure policy.
What you'll haveHigher-scoring proposals and cleaner manuscripts — more funding awarded and faster publication, the engine of an academic salary.
5
Keep the industry option open
Why this pays: Astro PhDs are prized in data science, quant finance, and ML research, where total comp often exceeds academia's top of the range. A public, AI-literate portfolio doubles your leverage — either for a high-paying pivot or as a credible outside option that raises your academic offers.
ClaudeGitHub CopilotHugging Face
1
Package a research method as a clean, documented GitHub repo and a short write-up, and share models or demos on Hugging Face so your ML skills are visible to industry, not buried in a journal.
2
Translate your research into industry terms.
Copy-paste this prompt
I'm an astrophysicist whose research used [methods: Bayesian inference, large-scale simulation, neural emulators, time-series]. Rewrite my experience as a data-science / ML-research resume: map each astro skill to an industry equivalent, quantify impact, and list the 5 portfolio projects that would make me competitive for [role].
Be accurate about what you actually built — interviews test the claims. Keep unpublished institutional data out of public repos.
What you'll haveA visible, AI-literate portfolio — a credible high-paying pivot or the outside option that lifts your academic comp.
Your 12-month sequence to the top of the range
How the plays above stack into a path from median pay toward the $296,740 tier.
Month 1
Wire NASA ADS plus Undermind/Elicit into your literature workflow and Copilot/Cursor into your code editor.
Months 2-3
Rebuild one analysis pipeline with AI assistance and add unit tests; measure the time saved.
Months 3-6
Prototype one ML surrogate or classifier for your data with proper uncertainty quantification.
Months 6-12
Use AI to sharpen a grant and a paper, and publish a clean method repo that doubles as an industry portfolio.
Next steps for an Astrophysicist
Some links below are affiliate or partner links. PayCrunch may earn a commission if you enroll or subscribe through them, at no extra cost to you. Wage figures on this page still come from the Bureau of Labor Statistics, not from these programs.
Astrophysicist work is specific enough that a stamped 'check out these courses' block would be noise. BLS files this work as Physicists (SOC 19-2012). O*NET Job Zone 5 is typical: graduate or professional school, so the honest next credential is a graduate-level or professional certificate — not a random catalog dump.
The occupation's listed knowledge areas include Physics and Engineering and Technology; the links search those subjects, not a generic 'career courses' list.
Astrophysicists in this dataset list Amazon Web Services AWS software among the tools in use, so a program that names that stack is a better fit than a survey course.
FlexJobs screens remote, hybrid, freelance, and flexible listings so you are not wading through unverified ads. This is a job-board search for Astrophysicist work, not a claim that they list a counted SOC 19-2012 inventory.
A an Astrophysicist resume you can submit beats a blank page. Resume Now is a resume builder — we are not claiming an occupation-specific template library for SOC 19-2012.
An Astrophysicist resume that names the actual tasks on this page beats a blank template when you apply.
What Astrophysicists earn by state
These are the Bureau of Labor Statistics’ own figures for Physicists, state by state — not a cost-of-living adjustment applied to the national number. Only states employing at least 500 people in the occupation are shown, because a state median drawn from a handful of workers is noise rather than a signal.
California
$207,490
highest of them · +20% vs the national median
Michigan
$94,520
lowest of the 11 states that qualify · -45% vs the national median
The same job pays $112,970 more a year at the median in California than in Michigan — 120% higher. That gap is what the Bureau measured, before any question of what it costs to live in either place. California also carries the top of this job’s range, $296,740 — the figure quoted at the head of this page.
Source: U.S. Bureau of Labor Statistics, Occupational Employment and Wage Statistics, May 2025, SOC 19-2012. 11 states clear the 500-employee reporting floor for this occupation; those below it are left out rather than shown with a wide error band.
Free data. Use any of it.
PayCrunch publishes verified, BLS-sourced salary + AI-playbook data on 1,000+ professions — free, no signup.
No — AI can't pose the physical question, judge whether a result is physically sensible, or take responsibility for a claim about the universe. It accelerates the search, the coding, and the drafting. The field's data deluge makes those who wield it well more valuable, not less; they publish more and build methods others rely on.
Can I trust an LLM to do the math or cite papers?
Not without checking. LLMs confidently produce wrong derivations and fabricated references. Re-derive equations yourself, verify every citation in NASA ADS or arXiv, and unit-test any code before it touches a result. Use AI to draft and accelerate, then verify like a referee would.
Is it worth learning deep learning if I'm a theorist?
Yes. Neural emulators, simulation-based inference, and ML classifiers are now core methods across theory and cosmology. The ability to build them is what makes a methods paper stand out — and what makes you employable far beyond academia if you choose.
How does AI actually raise an astrophysicist's pay?
Indirectly but powerfully: more papers and stronger proposals lead to grants, promotion, and senior or faculty positions; and the same ML skills open industry roles paying above academia's top of the range. AI is the multiplier on the output all of those rewards track.
Should I disclose AI use in papers and proposals?
Yes — follow each journal's and funder's policy, which increasingly require it. Never list AI as an author, never let it generate data or references, and keep a record of where you used it. Transparency protects both your integrity and your career.
Methodology & sources
Salary (median, 10th, top of the range) — U.S. Bureau of Labor Statistics, OEWS.
By state — the Bureau of Labor Statistics’ own state medians, limited to states employing at least 500 people in the occupation. No cost-of-living arithmetic is applied to a wage anywhere on this page.
The plays — PayCrunch's own step-by-step guidance using publicly available AI tools. Tool names/URLs are real and current as of August 2026; prompts written to work as-is. Verify any professional output before relying on it.