Pokee-AI/PokeeResearchOSS

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

1076 stars · 411 forks observed · 2026-08-28

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

An open-source repository for Pokee's 7B-parameter DeepResearch agent that performs multi-turn web search and content reading to answer complex questions with up-to-date online information. It includes inference code, tool integration, and benchmark evaluation across QA datasets like HLE, GAIA, and BrowseComp.

Use cases

• build an AI agent that researches questions using live web search
• run a small deep research model locally on a single GPU
• evaluate a research agent on QA benchmarks like GAIA and BrowseComp
• generate citation-rich research reports from web sources
• self-host an alternative to paid deep research APIs
• integrate web browsing tools into an LLM agent pipeline

When to choose

• you want an open 7B research agent you can run on a single 80GB GPU
• you need multi-turn web search with tool integration for complex question answering
• you want to benchmark deep research agents on standard QA datasets
• you prefer self-hosting over per-query API costs

When to avoid

• you need a fully managed service without GPU infrastructure
• your hardware has far less than 80GB GPU memory (untested configurations)
• you need a general chatbot rather than a web-research agent
• you want a model larger than 7B parameters for maximum accuracy

Facets

library · maturity active

agent-framework rag llm-inference search-engine machine-learning artificial-intelligence large-language-models python deep-research web-search-agent 7b-model qa-benchmarks tool-use ai-agents retrieval-augmented-generation search gpu docker linux

Member repositories

For agents

markdown · JSON · MCP: product_card(name="Pokee-AI/PokeeResearchOSS")