Student pricing on AI tools is real but uneven: some vendors give the paid plan away on proof of enrolment, others take twenty percent off, and several advertise a discount that turns out to be the ordinary annual rate. The six below have terms we checked against the vendor’s own page, with the date on each listing.
They are grouped by what a student day actually contains: reading sources, sitting in lectures, writing, and coding.
NotebookLM: reading a stack of papers without losing the citation
NotebookLM inverts how an assistant works. It answers only from the PDFs, web pages, YouTube videos and lecture recordings you put in the notebook, and every answer links to the exact passage it came from. For a literature review that has to be cited accurately, that constraint is the whole point.
Audio Overview turns a set of sources into a two-host discussion you can listen to on the way to campus, which is the closest thing to revision that does not feel like revision.
Student terms: the free tier allows 50 sources per notebook. Students in eligible countries get Google AI Pro free for 12 months, which raises that to 300 sources and five times the generation limits.
GitHub Copilot: free for verified students
GitHub Copilot is free for verified students, teachers and maintainers of popular open source projects, through the GitHub Student Developer Pack. That is the full $10 plan, not a trimmed version.
For a computer science degree it changes the pace of lab work: completions inside VS Code, JetBrains and Xcode, chat that understands the repository, and agent mode for multi-file changes.
Student terms: free with verification. Without it, $10 a month, and the free tier still gives 2,000 completions a month.
Gemini: the free tier that covers most coursework
Gemini has the widest free tier of any assistant. Deep Research, Canvas, Gems and image generation all work without paying, under lower usage caps, and it reads audio, video, images and PDFs in one thread. Recording a lecture and asking questions about it costs nothing.
Student terms: free tier is generous on its own. Google AI Pro at $19.99 is free for 12 months for students in eligible countries, adding 5 TB of storage that also solves the shared-drive problem.
Otter.ai: lectures transcribed while you listen
Otter joins a recorded class or runs on your phone in the room, transcribes as people speak, and produces a summary with the points that mattered. Searching across a semester of lectures for the one definition a professor gave in week three is the feature that earns it.
Student terms: 20 percent off Pro with a .edu address, bringing it to $13.59 monthly or $6.67 on annual billing. The free plan gives 300 minutes a month, which covers about four lectures.
Grammarly: corrections inside the document you are writing
Grammarly works where you write, in Google Docs, Word and the browser, rather than in a separate window. Beyond spelling it flags tone, which matters for the emails to supervisors that students routinely get wrong.
The free plan is permanent and covers spelling, grammar and 100 AI prompts a month. That is enough for most essays.
Student terms: education pricing on request. Pro is $12 a month on annual billing, $30 monthly, so the annual rate is the only one worth considering.
Julius AI: statistics coursework without writing Python
Julius takes a dataset and answers questions about it in plain language, writing and running real Python behind the conversation. The code stays visible, which means you can check the method and learn from it rather than trusting a number.
For a statistics or economics module, it removes the gap between knowing what test you need and knowing how to run it.
Student terms: students and educators get 50 percent off, bringing Plus to $10 a month. The free tier handles charts and small files.
What to check before you rely on any of them
Verification is the friction point. Most vendors want a school email, and some want a student ID through a verification service, so allow a day before a deadline rather than an hour.
The second thing is your institution’s policy on generated text. Transcription, explanation and code completion sit differently from submitted prose in almost every academic integrity code, and the rules are set by department rather than by the vendor.