Awesome list navigation
awesome-ai-scientists
A curated collection of resources for building “AI Scientist” systems — AI that assists scientific discovery across the full research lifecycle.
We map the tools, papers, datasets, frameworks, and workflows that help machines (and the humans who build them) do real science.
Start exploring
Three entry points into the navigation layer.
Browse by resource type
Open the right part of the list without scanning the whole README.
Papers
Foundations & surveys
Landmark systems, project papers, and field overviews.
Tools
Literature tools
RAG, search, metadata APIs, and scientific language models.
Tools
Planning & automation
Experiment design, optimization, and lab automation.
Benchmarks
Evaluation sets
Benchmarks for reasoning, coding, and agent performance.
Datasets
Scientific datasets
Core datasets for evaluation and training.
Models
Models & checkpoints
Hugging Face resources and domain-specific model entries.
Frameworks
Agent frameworks
General-purpose orchestration frameworks used by builders.
Meta
More lists
Adjacent awesome lists and broader resource hubs.
Browse by workflow or capability
Navigate the list the same way scientific agent systems are built and evaluated.
Review
Literature review
Search, retrieval, synthesis, and scientific context.
Ideation
Hypothesis generation
Systems for novelty, ideation, and research framing.
Planning
Experiment planning
Optimization, experimental design, and decision loops.
Execution
Tool use
Lab tooling, APIs, and systems that act through tools.
Agents
Multi-agent systems
Debate, coordination, and specialist-agent architectures.
Evaluation
Evaluation
Benchmarks, datasets, and agent capability measurement.
Reporting
Communication
Writing, reporting, figures, and manuscript tooling.
Safety
Safety & governance
Dual-use, policy, validation, and responsible AI resources.
Start here
Three fast entry points, depending on what you need from the list.
Current highlights
A quick scan of standout entries pulled from the current list.
Foundations
The AI Scientist v2
A flagship end-to-end system entry in the current list.
Literature
PaperQA2
A strong starting point for literature-grounded workflows.
Planning
BayBE
Bayesian optimization tooling for real experimental work.
Evaluation
ScienceAgentBench
A benchmark built around real scientific workflows.
Biology
AlphaFold 2/3
A central biology entry with broad downstream relevance.
Frameworks
LangGraph
A common orchestration layer for stateful agent systems.
Full awesome list