lupantech/chameleon-llm

Health v2 · maintenance only

Adoption not part of the score

1139 stars · 82 forks observed · 2026-08-28

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

Chameleon is a research framework for plug-and-play compositional reasoning with large language models, where an LLM planner synthesizes programs that compose tools such as vision models, web search, Python execution, and rule-based modules. It accompanies the NeurIPS 2023 paper and provides code and benchmarks (ScienceQA, TabMWP) demonstrating enhanced reasoning with GPT-4/ChatGPT.

Use cases

• build an llm agent that chains tools like search and python execution
• reproduce the chameleon compositional reasoning paper results
• evaluate gpt-4 on scienceqa and tabmwp benchmarks
• prototype a planner that composes modular tools for question answering
• augment llm reasoning with vision models and knowledge retrieval

When to choose

• you want a research-grade example of LLM tool-use planning and composition
• you need to reproduce or extend the Chameleon paper's benchmarks
• you're exploring plug-and-play tool augmentation for GPT-4/ChatGPT

When to avoid

• you need a production-ready, maintained agent framework
• you want to use open-source local LLMs rather than OpenAI APIs
• you need long-term support, as the repo is research code from 2023

Facets

framework · maturity maintenance

agent-framework llm-inference prompt-engineering rag nlp artificial-intelligence large-language-models python compositional-reasoning tool-use gpt-4 research-code planner ai-agents natural-language-processing

Member repositories

For agents

markdown · JSON · MCP: product_card(name="lupantech/chameleon-llm")