Literature Search
Search 400M+ scholarly works without leaving Academe, then pull relevant papers into your project.
How to search
Open the Literature panel from the left sidebar. Type a topic, a specific question, an author, or a string of keywords. Academe queries the scholarly index and ranks results by relevance.
Searches support the filters you’d expect: year range, open-access only, peer-review status, and source type (journal, preprint, book, thesis, dataset).
What you see for each result
Title, authors, and year
Abstract on hover
Peer-review status
Venue quality indicators
Open-access availability
Citation count
Importing results
Click the import button on any search result to add it to your project. If the paper is open-access, Academe downloads the full text automatically. Either way, the metadata lands in your library: title, authors, abstract, and DOI.
Once imported, Academe can search, cite, and use the paper in your conversations and writing.
Assistant-guided discovery
Beyond manual search, you can ask the Research Agent to find relevant papers for you. A few prompts that work well:
- “Find recent papers on X that I haven’t read yet.”Academe diffs your library against the corpus to surface new-to-you work.
- “Most-cited papers in this area from the last two years.”Popularity meets recency, bounded by your field.
- “Find papers that contradict my main argument.”The agent actively seeks counter-evidence and contrary findings.
Research digest
Enable the weekly research digest in your project settings. Academe searches for new papers relevant to your research and sends an email summary, including papers that may challenge or contradict existing findings.
Data sources
One unified index covers what researchers actually reach for:
Academic papers
Clinical trials
Bring your own data
Where search differs
- Beyond keywordsSemantic search finds papers that don't share keywords with your query but still address the underlying question. You don't need to guess the exact vocabulary of a subfield.
- Transparent screening criteriaNarrow a result set by age, method, population, outcome measure, or any custom criterion. Academe shows why a paper was kept or excluded for each criterion, not just overall.
- Scale to a thousand papersRun analysis against up to 1,000 results at once while keeping the candidate pool queryable.
- Project-aware rankingRanks against your project context, including existing papers, notes, and stated research focus, so relevance is tuned to you rather than generic popularity.
How search works under the hood
Four stages, each transparent if you want to look at them:
- 1. Specify the questionAcademe helps sharpen broad questions with suggested refinements: mechanism, dosage, onset, population, comparison, timeframe. The clearer the question, the better the retrieval.
- 2. Rank the corpus by similarityA custom embedding model scores every paper against the question. Features like citation count and recency feed into the ranking so recent high-signal work surfaces.
- 3. Deep review the top thousandLarge language models re-rank the top 1,000 candidates based on how closely their contents answer your question, not just whether they overlap in vocabulary.
- 4. Natural-language screeningOptionally run custom screening questions over the shortlist, including methodological filters, population thresholds, and exclusion rules, to get to a handful of papers worth reading in full.
Example searches
- Quick look-up"Online vs. face-to-face CBT for treating depression" - returns comparative studies with methods and outcomes you can inspect.
- Inside a research report"Biomarkers used for seizures associated with Lennox-Gastaut syndrome" structures a full report from the retrieved evidence.
- Systematic review shortlist"Targets for metastatic neuroendocrine prostate cancer" feeds the Gather step of a systematic review with pre-scored candidates.
Workflows that start from search
- Research reportsGenerate a structured 10+page literature overview from the papers you find.
- Systematic reviewsUse search as a seed to populate the Gather step of a full systematic review protocol.
- Research alertsSave a promising search as an alert so new papers matching it surface in your weekly digest.
FAQ
- Can I search for papers in languages other than English?Yes. Semantic search works across major research languages (English, Spanish, Portuguese, French, German, Chinese, Japanese, Korean, Arabic, and more). Results from non-English sources are returned in their original language with on-demand translation.
- Do you have access to paywalled papers or only open access?We surface the open-access version first when available. For paywalled papers, connect your institutional access and Academe will route requests through it where your license permits.
- How do you rank sources or determine quality?Ranking blends semantic relevance to your query, citation weight, recency, and peer-review status. Venue quality indicators surface alongside each result so you can judge for yourself.
- How many databases do you search?A single unified index. You do not have to assemble queries across databases by hand; one query hits academic papers, conference proceedings, preprints, and clinical trials.
- How often do you update the database?Academic papers update weekly. Clinical trial registry entries update on a regular schedule.