Course catalogue data with the ratings attached
The Coursera API returns online-course data as clean JSON.
5 active endpoints, on 0 and 1 credit tiers.
- POST/coursera/v1/search
- POST/coursera/v1/detail
- POST/coursera/v1/reviews
- POST/coursera/v1/partners
- POST/coursera/v1/subjects
What Coursera endpoints does ReefAPI ship?
5 live read endpoints. Read-only data API: no writes, no account actions, no dashboard access on the target site.
Coursera API
3 of 5 endpoints, ready to run
Courses and specialisations matching a query, each with its rating, review count, difficulty, duration, the partners behind it and the skills it teaches.
{ "ok": true, "meta": { "api": "coursera", "endpoint": "search", "mode": "live", "latency_ms": 980.2, "record_count": 12, "cache_hit": false }, "data": { "results": [ { "id": "nOOfCDWeEeuiZgo2K4rorQ", "name": "Machine Learning", "url": "https://www.coursera.org/specializations/machine-learning-introduction", "slug": "machine-learning-introduction", "type": "SPECIALIZATION", "difficulty": "BEGINNER", "duration": "ONE_TO_THREE_MONTHS", "rating": 4.89, "review_count": 39269, "partners": [ "DeepLearning.AI", "Stanford Online" ], "skills": [ "Unsupervised Learning", "Supervised Learning", "Model Training" ], "is_free": false, "is_coursera_plus": false, "image": "https://d15cw65ipctsrr.cloudfront.net/3a/9d2a7af297483a845340bcfbac6f1e/MLS.course-banners-01_Course-Logo-.png", "tagline": "#BreakIntoAI with Machine Learning Specialization\nOffered by DeepLearning.AI, and Stanford Online" }, { "id": "8UjeMk-mEeit4g4GsxE4dg", "name": "Machine Learning with Python", "url": "https://www.coursera.org/learn/machine-learning-with-python", "slug": "machine-learning-with-python", "type": "COURSE", "difficulty": "INTERMEDIATE", "duration": "ONE_TO_THREE_MONTHS", "rating": 4.67, "review_count": 18454, "partners": [ "IBM" ], "skills": [ "Unsupervised Learning", "Supervised Learning", "Model Evaluation" ], "is_free": false, "is_coursera_plus": true, "image": "https://s3.amazonaws.com/coursera-course-photos/3c/77927732934dc0a10a44bd3306833d/1200x1200px_1001823008.jpg", "tagline": "Offered by IBM" }, { "id": "Q0Fc_Yl0EeqdTApgQ4tM7Q", "name": "IBM Machine Learning", "url": "https://www.coursera.org/professional-certificates/ibm-machine-learning", "slug": "ibm-machine-learning", "type": "PROFESSIONAL_CERTIFICATE", "difficulty": "INTERMEDIATE", "duration": "THREE_TO_SIX_MONTHS", "rating": 4.61, "review_count": 3700, "partners": [ "IBM" ], "skills": [ "Unsupervised Learning", "Exploratory Data Analysis", "Autoencoders" ], "is_free": false, "is_coursera_plus": true, "image": "https://d15cw65ipctsrr.cloudfront.net/78/0b8c921b6346f69278f39ba6ca8128/Professional-Certificate---IBM-Machine-Language.png", "tagline": "Prepare for a career in machine learning\nOffered by IBM" } ] } }
How the Coursera API works
Coursera is a normal ReefAPI surface — the same four rules that hold for every other engine on the key.
No OAuth app, no request signing, no per-site account. One key covers all 184 engines.
Every route is a POST with a JSON body. Parameters are validated against the published schema before anything is charged.
Credits, not seats. Failed and blocked calls are never charged, and cache hits cost nothing.
One envelope everywhere. meta carries latency_ms, record_count and the endpoint that answered.
Comparing what is on offer for one skill
Course marketplaces are optimised to sell you the next course, not to let you compare. The comparison needs rating, volume, level and time commitment side by side.
{"query": "machine learning", "level": "Beginner"}Every row carries the rating, the review count, the difficulty, the duration band, the partners and the skills — enough to rank without a second call.
{"slug": "machine-learning-introduction"}Then the written reviews for the shortlist, where the actual complaints live.
The skills list on each row is the underrated field: it is how you tell two courses with the same title and rating apart.
curl -X POST https://api.reefapi.com/coursera/v1/search \
-H "x-api-key: $REEF_KEY" \
-H "content-type: application/json" \
-d '{"query":"machine learning"}'{
"ok": true,
"data": { … },
"meta": {
"api": "coursera",
"endpoint": "search",
"mode": "live",
"latency_ms": …,
"record_count": …
},
"error": null
}Coursera ids, slugs, and the enum casing that differs between input and output
Two things break integrations here. The id you get back is not an id you can send anywhere, and the vocabulary you filter with is not the vocabulary you receive. Everything below was read live on 2026-08-27 from search for "machine learning", detail for slug machine-learning, and detail for the parent specialization.
| Field | Measured value | Note |
|---|---|---|
| id | nOOfCDWeEeuiZgo2K4rorQ | A 22-character opaque id. No action accepts it as input; every action takes slug or url |
| slug | machine-learning | The last path segment of the URL, and the real key. It is not the title: slug machine-learning returns "Supervised Machine Learning: Regression and Classification" |
| type (returned) | COURSE, SPECIALIZATION | Upper snake case |
| type (search filter) | Courses, Specializations, Professional Certificates, Guided Projects, Degrees | Title case and plural. The filter vocabulary and the returned vocabulary do not match |
| difficulty | BEGINNER, INTERMEDIATE, MIXED returned; Beginner, Intermediate, Advanced, Mixed accepted | Same split as type |
| duration | ONE_TO_FOUR_WEEKS, ONE_TO_THREE_MONTHS, THREE_TO_SIX_MONTHS | A bucket, not a number of hours. Returned only, not filterable |
| price | Not returned by any action | Only the booleans is_free and is_coursera_plus tell you anything about cost |
| ratings | average 4.89, rating_count 32,841, comment_count 6,187 | A five-point scale. ratings_by_star is keyed with words: five, four, three, two, one |
| total_enrollments | 1,237,972 on the course, 830,254 on the specialization that contains it | A course and its parent program count separately. Never add them up |
| review author | "RG", "AC", "YD" | Initials only. Coursera publishes no reviewer name, avatar or profile link |
search returns 12 results per page, and on a live call meta.total_count, meta.total_pages and meta.next_page all came back null with has_more false on page 1, yet page 2 returned 12 more results. Do not treat has_more as a stop signal here; page until a page comes back empty, up to the parameter ceiling of 80. Note also that what_you_will_learn and recommended_background arrive with HTML entities left in place, as in "NumPy & scikit-learn", so unescape before you display them.
What a catalogue row carries
Measured across searches and detail lookups.
A 4.9 from thirty-nine thousand reviews and a 4.9 from eleven are not the same recommendation, and both numbers arrive on the search result rather than behind a detail call.
A specialisation is a sequence of courses and the row says so with its type. Ranking a multi-month sequence against a single course as if they were comparable is the most common mistake in this data, and the type field is what prevents it.
Rows carry the institutions behind the course. On this platform the partner is most of the perceived value, and it is a first-class field rather than something to parse out of a title.
Each course lists the skills it teaches. That is the axis on which two similarly-rated courses actually differ, and it is the one a recommendation should be built on.
The catalogue exposes ratings, structure and content rather than the price a given learner would pay, which varies by subscription, region and promotion. Treat this as course data, not pricing data.
What people build with Coursera
The jobs this data is most often used for.
endpoints
credits per call
Course-aggregators call search to list Coursera courses by subject, level and rating.
Learning apps use detail and reviews to show course info and learner feedback.
Skills-mapping tools use skills and subjects to match courses to a career path.
What Coursera data costs
The cheapest call here is 0 credits, so $15/mo (Pro) buys 10,000 of them — $1.50 per 1,000 credits. Credits roll over and never expire, and failed or blocked calls are not charged.
Full pricing →- 1,000 free credits on signup, no card
- One key, all 184 APIs, one credit pool
- Failed and blocked calls are never charged
- Credits roll over and never expire
Call it in two lines
Sign up, get 1,000 credits and one key that works on every engine. Then this is the whole protocol.
curl -X POST https://api.reefapi.com/coursera/v1/search \
-H "x-api-key: $REEF_KEY" \
-H "content-type: application/json" \
-d '{"query":"machine learning"}'import requests
r = requests.post(
"https://api.reefapi.com/coursera/v1/search",
headers={"x-api-key": REEF_KEY},
json={
"query": "machine learning"
},
)
print(r.json()["data"])Have a question? We got answers.
The questions people actually ask before wiring up Coursera.
Get a free key →Does the Coursera API return course prices?▾
No. Across search, detail and reviews there is no price, currency or discount field of any kind. What you get is is_free and is_coursera_plus on a search row, which tell you whether the item can be taken for nothing and whether it is included in a Coursera Plus subscription. This differs from the Udemy engine, where detail.price is the list price and the pricing block holds the real one; here there is no number to disagree about because Coursera does not publish one on the page.
Why does slug machine-learning return a course with a different name?▾
Because slugs are permanent and titles are not. On 2026-08-27, detail for slug machine-learning returned name "Supervised Machine Learning: Regression and Classification", the current occupant of Andrew Ng's original ML course URL. Key your records on slug, keep name as a display field, and expect a name to change under a stable slug.
How do I tell a course from a specialization, and how do I fetch each?▾
A search row's type comes back as COURSE or SPECIALIZATION and its url shows the same thing: /learn/<slug> for a course, /specializations/<slug> for a program. detail auto-detects when you pass url. If you pass a bare slug it assumes a course, so add type "specialization" for a program. A specialization's detail adds courses[] with the slug of each course inside it plus course_count, which was 3 for machine-learning-introduction.
What does detail return that search does not?▾
Rather a lot. A live detail read added average_instructor_rating 4.95, content_satisfaction_score 98.0, page_views_last_month 1,863,125, launched_at 2022-06-14T21:27:57.322Z, certificates ["SPECIALIZATION","VERIFIED_CERTIFICATE"], 32 subtitle_languages against a single primary language, the ratings_by_star breakdown, and instructors with courses_taught and learners_reached per person (Andrew Ng: 51 courses, 9,929,504 learners). page_views_last_month in particular is a demand signal you will not find on a search row.
Why do rating_count and comment_count differ so much?▾
Because most learners rate without writing. The measured course returned rating_count 32,841 against comment_count 6,187, so roughly one rating in five carries text. ratings.average is computed over the ratings, not over the comments, which is why it will not match an average you calculate from the review text you can fetch.
How many reviews can I pull, and who wrote them?▾
Twenty, and effectively nobody you can identify. A live reviews call returned 20 rows, each with review_id, author, rating and date, and Coursera exposes no deeper pagination without a login. The author field is initials only, "RG" and "AC" in the measured rows. review_id is a composite, 68750271~COURSE!~T7hf5jWeEeuGCBL8_hyTUQ, with the course's opaque id after the second tilde. date is day precision, midnight Z. meta.review_count and meta.rating_average came back null on the reviews call, so take the totals from detail's ratings block.
Why do my filter values not appear in the response?▾
Because the filters take human-readable values and the response returns Coursera's internal enums. You send level "Beginner" and type "Courses"; you get back difficulty "BEGINNER" and type "COURSE". Both filters are declared on_invalid=ignore, so a wrong value silently widens the search instead of erroring. Map in both directions rather than comparing the strings directly.
Should I add a specialization's enrollments to its courses'?▾
No, they overlap. detail for the course machine-learning reported total_enrollments 1,237,972 while its parent specialization machine-learning-introduction reported 830,254. Each number counts enrollments in that product, and someone taking the specialization is also enrolled in its three courses. Report them separately or pick one level to aggregate at.
What is the Coursera API?▾
Coursera API is a ReefAPI endpoint group for coursera It returns live JSON through POST requests under /coursera/v1.
Is the Coursera API free to try?▾
Yes. ReefAPI starts with 1,000 free credits, no card required. Coursera calls use the same shared credit balance as every other ReefAPI engine.
Do I need a Coursera login or account?▾
No login to Coursera is needed for the API response. You call ReefAPI with your x-api-key header, and the playground can run live examples before you create a production key.
How fresh is the Coursera data?▾
The page example is captured from a live search call, and production requests fetch live data through ReefAPI rather than a static sample.
How many credits does the Coursera API use?▾
Coursera actions currently cost 1 credit per successful call. Failed or blocked calls are free, and all APIs draw from one credit pool.
Can I call Coursera from an AI assistant or MCP client?▾
Yes. Connect ReefAPI once through MCP and your assistant can call coursera actions with the same key, credit pool and JSON envelope used by normal REST requests.
25 Media, Film & Knowledge APIs on the same key
One key, one credit pool, one response envelope. If you are pulling Coursera, you are one call away from the rest of the category — no second contract, no second integration.
Need something this API does not do?
Name the endpoint, the field, or a source we do not carry yet. We ship new APIs every week and you would be first to get the key. Real people read every message and reply the same day.
Try it on your own data before you pay anything
The call above is the real endpoint, not a recording. A free key gives you 1,000 credits, the other 183 APIs, and the same envelope everywhere.
Endpoints, parameters and credit costs on this page are read from the live catalog and cannot drift from what the API accepts. Field notes were captured on 2026-08-30.