litsurvey
v{{VERSION}} · Scientific Literature Survey
Run
History
Docs
Report a bug
Search & lookup (no LLM involved)
Search
keyword search across seven indexes (OpenAlex, Semantic Scholar, arXiv, PubMed, preprint portals)
Paper
one paper's full record: abstract, DOI, ids, BibTeX
Cites
papers that cite a given paper
Refs
papers a given paper cites
Related
similar papers (recommendation engine)
Open access
legal free copy of a DOI (Unpaywall)
LLM agent (needs a backend)
Novelty
prior-art verdict for a claim
Research
cited state-of-the-art survey for a question
Where does the LLM run?
Subscription CLI
Claude Code, Codex or Gemini CLI, signed in with your Claude Pro, ChatGPT or Google plan. Your text goes to that vendor.
Local model
Ollama on this machine. Nothing leaves it.
Cloud API key
OpenAI-compatible or Anthropic API. Your text goes to the provider.
CLI tool
Provider
openai-compatible
anthropic
API base URL
Model
Max rounds
Keyword query
Run
Max results
From year
Output file (optional; extension picks format: .bib .ris .csv .json .md)
Past runs
stored in ~/.litsurvey/runs
time
mode
input
results