
by FutureHouse
PaperQA2 is a Python package for retrieval augmented question answering over scientific papers and other documents. It gives answers with in-text citations and can run as an agent that refines its own searches.
AgentFree
Step 1: Install the package with pip install paper-qa.
Step 2: Put the PDFs you want to query in a folder.
Step 3: Run pqa ask with your question from that folder.
Step 4: Read the cited answer, or use the library to add and query documents in code.
Step 5: Change the LLM, embedding model or settings to fit your setup.
PDFs, text files, Microsoft Office documents and source code files.
It supports all LiteLLM providers, and it can also run with locally hosted models and local embedding models.
The agent can search for papers, gather evidence and write an answer, and it can refine queries and answers over several steps.
It fetches metadata such as citation counts and journal quality from multiple providers, and it relies on services such as Semantic Scholar and Crossref.
FutureHouse PlatformScientific-agent platform for literature search, deep research, evidence synthesis, novelty checking, and chemistry and biology research workflows
AviaryAviary is an open-source gymnasium from FutureHouse for defining custom language-agent reinforcement learning environments. It ships with environments for math, general knowledge, biological sequences and scientific literature search.