Five formats, from free to $13,274. Free university programmes are the best option and are mostly closed to non-US students. Accredited online research costs $7,465 and buys transferable college credit. Online 1:1 mentorship runs £2,000–£5,000. Residential summer schools cost £6,995–£9,995 and are mostly buying you a fortnight in a university city. The right answer depends on four things about your student, not on which programme markets hardest.
Most comparison pages in this market rank programmes as though one is simply better than the others. They are not comparable in that way, any more than a bicycle is better than a boat.
What actually varies is the format, and formats suit different students. Below is what each one is, what it costs, and the specific situation it is right for.
We run one of these. Our bias is obvious and we have tried to be useful anyway — including by telling you, several times, to go elsewhere.
The five formats
| Format | Examples | Cost | Length |
|---|---|---|---|
| Free university programme | RSI, MIT Beaver Works, Stanford SIMR | Free (SIMR: $50 application fee) | 4–6 weeks residential |
| Paid university programme | UCSB Research Mentorship Program | $5,675 commuter · $13,274 residential | 6.5 weeks |
| Accredited online research | Pioneer Research Institute | $7,465 + up to 2% materials | Online, 4 college credits |
| Online 1:1 mentorship | OCAR, Immerse, Polygence | £2,000–£4,995 / from ~$3,000 | 10 hours to 20 weeks |
| Residential summer school | Immerse, Oxford Scholastica, Atlas | £6,995–£9,995 | 2 weeks |
All figures from each provider's own site, checked September 2026.
Start here: the free tier
If your student is eligible and competitive, apply to these before paying anyone. They are more selective than the universities their alumni attend, and nothing you can buy is better.
Research Science Institute (RSI) — free, about 100 students a year at MIT: one week of intensive classes, five weeks of research, a final week of reports. Year 12 equivalent only.
MIT Beaver Works Summer Institute — free for qualifying families, $2,400 for families with income and assets of $200,000 or more. Four weeks in July.
Stanford SIMR — a $50 application fee, waived below $80,000 family income, and no cost if selected. Around 50 students. Selection "will strongly favor local students."
Now the catch, stated plainly because most comparison pages bury it: all three are restricted to students living in and attending school in the United States. For an international applicant, the free tier is largely unavailable — which is the single biggest reason paid programmes exist at all.
UK students have their own free tier, which is genuinely good and almost never mentioned in these comparisons. We cover it in free research opportunities in the UK.
Four questions that decide the rest
Rather than ranking programmes, answer these.
1. Does your student need US college credit?
If yes, this narrows quickly. Pioneer Research Institute charges $7,465 and includes four transferable college credits through partnerships with Oberlin College and UNC–Chapel Hill, with every mentor a professor. That is a real, externally validated thing and expensive to provide.
If your student is applying through UCAS, college credit does almost nothing. UK universities make grade-based offers and do not award advanced standing for it.
2. Are you buying an experience or a document?
This is the question most families get wrong, because both are sold in the same language.
A residential summer school — Immerse at £7,495, Oxford Scholastica at £7,495, Atlas from £6,995 — buys two weeks in a university city with accommodation, meals, excursions and a certificate. That is a genuine experience and can be formative. What it rarely produces is a piece of work a tutor can interrogate.
A research mentorship buys a document: a question, a method, an analysis, a paper. What it does not buy is a summer to remember.
Neither is wrong. But £7,000 spent on the wrong one is an expensive mistake, and the marketing for both uses the words "research" and "Oxford."
3. How old is your student?
| Year group | What usually works |
|---|---|
| Year 9–11 | Subject exploration. Residential summer schools take students from 13; most research mentorships should not |
| Year 12 | The window for everything. Free programmes, mentorship, olympiads, EPQ |
| Year 13 | Too late for most programmes to affect the application — focus on the statement and the interview |
Independent research works once a student has enough subject grounding to frame a question that is genuinely theirs. Before that, the mentor ends up designing the project, which defeats the purpose and is visible at interview.
4. How many hours are you actually buying?
The comparator nobody presents. Divide before concluding anything.
| Programme | Price | Contact hours | Per hour |
|---|---|---|---|
| OCAR Oxford Scholar Programme | £2,000 | 10 | £200 |
| Immerse Publication Research | £4,995 | 19 | ~£263 |
| Immerse Classic | £2,695 | 10 | ~£270 |
| Polygence Core | from ~$3,000 | 10 | ~$300 |
| Pioneer Research Institute | $7,465 | Not published | — |
Most of this market prices 1:1 academic mentorship between roughly £200 and £300 an hour. Where a total is much higher, ask what the premium buys — usually accreditation, mentor seniority, duration or accommodation. Any of those may be worth it. If nobody can tell you which, that is your answer.
What none of them publish
One pattern is worth more than any individual comparison: almost nobody in this market publishes an acceptance rate.
Not RSI, not Beaver Works, not SIMR, not UCSB. Not Pioneer, not Polygence, not Immerse. Selectivity is the number that tells you whether a programme's name means anything on an application, and it is the number the entire industry omits.
We publish ours — about 60%, self-reported and unaudited like any figure in this market would be. That is not selective and we do not pretend it is. What we screen for is whether a student has the English, the subject grounding and the four to six hours a week to finish.
If selectivity is what you are buying, buy it from the free programmes, where it is real and enormous.
Where OCAR fits, and who should not choose us
We sit in the online 1:1 mentorship row, at the lower end of it on price.
The Oxford Scholar Programme is £2,000 over ten to fourteen weeks, ending in a 4,000–8,000 word paper. The Oxford Publication Fellowship is £3,400 over sixteen to twenty weeks and adds submission support. All our mentors are PhD candidates at Oxford, Cambridge, Ivy League or other leading UK universities. Payment plans are interest-free, which is worth noting against Polygence's stated 8% surcharge for paying over six months. Partial need-based scholarships exist — ask.
Choose something else if:
- your student is eligible and competitive for RSI, Beaver Works or SIMR — those are free and better;
- you need US college credit, where Pioneer's accreditation is a real advantage we do not offer;
- you want a residential experience, which we do not provide at any price;
- your student is under Year 12 and has not yet done enough subject work to own a question;
- you want external accreditation, where Immerse's British Accreditation Council status is something we do not have.
The decision, in one table
| If your student… | Look at |
|---|---|
| Lives in the US and is academically outstanding | RSI, Beaver Works, SIMR — apply free, first |
| Lives in the UK and money is tight | UNIQ, Sutton Trust, In2STEM, research placements, EPQ |
| Needs US college credit | Pioneer Research Institute |
| Wants two weeks on a campus | Immerse, Oxford Scholastica, Atlas |
| Is 13–15 and exploring | Residential summer schools |
| Is in Year 12 and wants a paper | Online 1:1 mentorship |
| Wants a journal submission | A mentorship whose provider names its target journals |
| Has the appetite but no research access | Online 1:1 mentorship |
That last row is the honest description of who paid mentorship is for. It is a common situation, especially for international students locked out of the free tier. It is not everyone's.
Frequently asked questions
Which high school research programme is best?
There is no single best one, because the formats differ more than the providers do. Free university programmes are the strongest thing available if your student is eligible and competitive. Beyond that, the right choice depends on whether you need college credit, a campus experience, or a finished piece of work.
How much should I expect to pay?
Free to $13,274. Online 1:1 mentorship clusters between £2,000 and £5,000; accredited online research is $7,465; two-week residential summer schools run £6,995–£9,995. Our full cost comparison sets out every published figure side by side.
Do colleges know which programmes are paid?
Increasingly, yes. Competitive admissions offices see thousands of applications and the well-known programme names recur. Write about the work rather than the brand — the ProPublica investigation into this industry is worth reading before you spend anything.
Is a summer school as good as a research project?
For a UK application, usually not, because Question 3 rewards something you can be questioned about in depth. For subject exploration at 14, a summer school is better. They answer different needs.
What if we cannot afford any of this?
Then use the free tier, which for UK students is genuinely strong, and an EPQ, which your school may already offer. Oxford's guidance is that tutors want to see how deeply you have engaged with your subject. It says nothing about how much the engagement cost.
How do I check a programme is what it claims?
Ask three questions: what is your acceptance rate, who awards any certificate, and what does my child walk away with if nothing gets published. Vague answers to any of the three tell you what you need to know.