RUC-NLPIR/Search-o1

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

1247 stars · 110 forks observed · 2026-09-27

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Search-o1 is a research framework that enhances large reasoning models (like QwQ and R1) with agentic search and retrieval-augmented generation during multi-step reasoning. It accompanies an EMNLP 2025 paper and provides code for reproducing experiments on benchmarks such as AIME, AMC, GPQA, and LiveCode.

Use cases

• add web search to a reasoning LLM's chain of thought
• run RAG-enhanced inference with QwQ or DeepSeek R1
• evaluate reasoning models on math and science benchmarks
• reproduce the Search-o1 paper experiments
• build an agentic search pipeline for LLM reasoning

When to choose

• you want to augment reasoning models with retrieval during inference
• you need a research baseline for search-enhanced LLM reasoning
• you're benchmarking on AIME, AMC, GPQA, or LiveCode

When to avoid

• you need a production-ready RAG service
• you don't have GPU resources for large reasoning models
• you want a plug-and-play chatbot rather than a research codebase

Facets

library · maturity active

rag agent-framework llm-inference search-engine machine-learning large-language-models python reasoning-models agentic-search research-framework emnlp-2025 math-reasoning retrieval-augmented-generation ai-agents natural-language-processing linux gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="RUC-NLPIR/Search-o1")