FutureHouse Platform
A suite of AI agents built to do real scientific research.
About
FutureHouse is a nonprofit lab (backed by Eric Schmidt) building an "AI Scientist". Its platform, launched May 2025, offers specialized agents: Crow for literature Q&A (built on the open-source PaperQA), Falcon for deep literature reviews with access to databases like OpenTargets, Owl for checking whether an experiment has been done before, Phoenix for chemistry experiment planning (built on ChemCrow), and Finch for data-driven biology discovery.
FutureHouse reports its agents beat PhD-level researchers on literature retrieval precision in head-to-head tests (around 90% vs 67% on LitQA science questions). The platform launched with free access via web and API. Note: FutureHouse has said its AI systems are now part of Edison Scientific's platform, so check current access terms.
What it can do
- Literature Q&A with citations (Crow)
- Deep multi-source literature reviews (Falcon)
- Prior-work and novelty checks (Owl)
- Chemistry experiment planning (Phoenix)
- Biological data analysis (Finch)
- API access
Setup
- Go to futurehouse.org and sign up for the platform.
- Pick an agent: Crow for literature questions, Falcon for deep reviews, Owl for prior-work checks, Phoenix for chemistry planning.
- Enter your research question or task.
- Review cited answers and iterate.
- Use the API for programmatic access.
Prerequisites
Account signup; check current access and pricing terms.
Use cases
Works with
Trust score — 7/10
Scored 7/10 (metrics tier, metrics-v1). Evidence: growing but early-stage user base; high-profile backing; actively developed nonprofit lab; platform launched May 2025; open-source PaperQA/ChemCrow roots; platform terms shifting toward Edison Scientific. Manual review: LitQA results are impressive but provider-reported; Phoenix chemistry agent is explicitly experimental. License: Platform terms via futurehouse.org / Edison Scientific; underlying agent code open source..
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 6.0/10
- Maintenance · recency, releases, issue load
- 8.0/10
- Openness · license + docs
- 7.0/10
- Manual review · sanity read + red-flag check
- 7.0/10
- · growing but early-stage user base; high-profile backing
- · actively developed nonprofit lab; platform launched May 2025
- · open-source PaperQA/ChemCrow roots; platform terms shifting toward Edison Scientific
- · manual review: LitQA results are impressive but provider-reported; Phoenix chemistry agent is explicitly experimental
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.