From $100 per month · Free trial🇸🇬
AI platform for patent search, patent drafting and innovation intelligence with data on more than 214 million patents, from Singapore.
PaperQA2 overview and key facts (October 2026)
Open-source tool from FutureHouse that answers questions from scientific PDFs with citations and finds contradictions between papers.
Free
What is PaperQA2?
PaperQA2 is an open-source Python package from FutureHouse for retrieval-augmented question answering over scientific documents. It answers questions from local PDFs and text files with in-text citations and suits research teams with technical staff who want to run literature analysis on their own document collections.
PaperQA2 indexes local PDFs, text files and code, runs full-text search, fetches metadata such as citation counts from Crossref and Semantic Scholar and answers questions with citations. It handles text, images, tables and equations and supports every model available through LiteLLM, including local models via Ollama. It installs with pip and needs Python 3.11 or newer. The repository is Apache 2.0 licensed, had about 9,300 stars and was last pushed on September 25, 2026.
PaperQA2 AI features and use cases
- Answers from your own papersPaperQA2 retrieves relevant passages from a local collection and writes an answer with in-text citations. Researchers query internal study archives with traceable answers.
- Finds contradictionsThe tool can identify conflicting claims across papers. Review teams spot disputed findings before they end up in a summary.
- Refines searches with agentsAn agentic workflow reformulates the query step by step and re-ranks passages with the language model. Answers improve on complex questions without manual search iterations.
What sets PaperQA2 apart
- Runs on your own infrastructure and can use local models, so documents need not leave the company.
- Apache 2.0 license allows commercial use.
Where PaperQA2 is strongest
Available internationally
Open-source library without a country focus, maintained by FutureHouse.
Questions to ask before you buy PaperQA2
This is a developer library, not a hosted service: installation, updates and model configuration are your team's job. When using commercial models, costs arise per request and document passages are sent to the model provider, so choose providers and settings accordingly. Check the licenses of the papers you index.
PaperQA2 reviews
No reviews yet. Be the first to share how your team uses it.
Every review shows its verification level. We read each one before it goes live. How we check reviews
Questions about PaperQA2
What does PaperQA2 cost?
The software is free under the Apache 2.0 license. You pay only for the language models you connect.
Which models does it support?
All models available through LiteLLM, including OpenAI, Claude, Gemini and local models via Ollama or llamafile.
Is the project maintained?
Yes. On September 28, 2026 the repository had about 9,300 stars and its last push was on September 25, 2026.
PaperQA2 update historyno updates yet
No changes recorded for PaperQA2 yet.
Changes the vendor made and confirmed user reports, each checked against the vendor's own pages. We review every tool monthly.
Alternatives to PaperQA2
Tools in the same category (Literature research) with the most overlapping markets.
From $14 per user/month · Free plan🇺🇸
Research tool with a browser extension that shows whether studies support or contradict a claim, plus an AI assistant with sources.
From $49 per user/monthAvailable internationally
AI research across more than 138 million papers with research reports, systematic literature reviews and data extraction.