Free during beta • No credit card required

Find research themes and candidate gaps in your literature, fast, with citations and confidence.

Upload 10–75 papers. LitRadar maps your corpus, clusters themes, highlights under-covered intersections, and generates evidence-linked gap hypotheses with stability scores.

Candidate gaps are relative to the uploaded corpus.

Evidence-linkedRepeatablePrivate by design
Theme map
Navigate clusters & representative excerpts
Candidate gaps
Coverage + missing connections
Stability score
Confidence labels for conclusions
LitRadar Report
Stability: 0.78
Sections
Theme Map
Candidate Gaps
Evidence
Repeatability
Corpus
42 papers • 1,286 chunks
Manifest: lit-9f3a…
Theme Map (Visualization)
Clusters shown for navigation, not distance claims.
Candidate gaps

Relative to the uploaded corpus.

Under-studied intersections
Theme B × Population X
High
Weak bridge between themes
Theme A ↔ Theme D
Medium
Sparse coverage region
Few nearby excerpts
Exploratory
Evidence links included in the full report.
Evidence-linked summaries
Deterministic processing + stability checks

Preview is illustrative. Reports cite the exact excerpts used.

Built for real research workflows

A literature review cockpit, not a black box.

LitRadar accelerates exploration and hypothesis generation, then points you back to the evidence so you can verify and write with confidence.

Undergrads

Quickly orient to a topic and learn how themes connect without keyword tunnel vision.

Grad / PhD

Compress the “scan phase” from weeks to hours, then validate gaps with evidence-linked outputs.

Researchers

Compare corpora, check coverage, and document reproducible analysis choices.

A repeatable pipeline

How LitRadar works

Designed to be auditable: the same inputs produce the same outputs, with manifests and stability checks.

1) Upload papers
Recommended 30–75 for best coverage. 10–20 works for fast exploration.
2) Extract + normalize
Consistent parsing and stable chunk IDs make results reproducible.
3) Embed + analyze
Cluster themes, measure coverage, and flag weak connections between topics.
4) Report with citations
Evidence-linked summaries and confidence labels you can defend.
Pro tip: Use larger corpora for stronger claims. Small sets surface ideas faster, but results should be treated as exploratory.

What you get

Analysis outputs students can cite

Every conclusion links back to the excerpts that support it, with confidence labels to keep claims honest.

Theme Map (Visualization)
Navigate clusters and representative excerpts. Visualization is for exploration.
Readable in seconds
Theme Clusters
Grouped themes across your corpus, with coverage and representative passages.
Cluster summaries + evidence
Similarity Matrix
See which papers are close, which are distant, and which ones bridge themes.
Cross-paper relationships
Evidence-linked Summaries
Summaries that cite the exact excerpts used, so you can verify quickly.
Cites every claim
Candidate Gaps (Corpus-relative)
Under-covered intersections and weak connections, labeled by confidence.
High / Medium / Exploratory
Coverage & Intersections
Identify what’s well-covered and what’s thin across methods, populations, or measures.
Find “missing combos”

Trust and reproducibility

Repeatability you can audit

LitRadar is designed to produce stable results for the same inputs and to clearly signal when inputs change.

Built like a methods section

The report includes a corpus manifest, deterministic processing settings, and stability signals so you can trust what you’re seeing and explain it to an advisor.

Manifest: lit-9f3a…
Corpus manifest
File hashes + chunk counts confirm exactly what was analyzed.
Deterministic processing
Stable chunk IDs + fixed seeds keep reruns consistent.
Stability score
Gaps are ranked by how consistently they appear across runs.
Important: If your corpus changes, your results should change. LitRadar tells you when that happens.

Your data stays private

No training on your PDFs
Your documents are not used to train models.
You control deletion
Delete projects when you’re done.
Store only outputs
Keep reports and visuals you choose to save.

Beta Terms: By using LitRadar during beta, you agree to our privacy policy. We may use anonymized, aggregated statistics to improve the service.

Ready to map your literature?

Start with 10–20 papers to explore quickly, then scale up for stronger, more stable conclusions.