# microsoft/JARVIS

JARVIS, a system to connect LLMs with ML community. Paper: https://arxiv.org/pdf/2303.17580.pdf

Repository: https://github.com/microsoft/JARVIS
Canonical: https://ross.abutalabs.com/products/jarvis
Language: Python
License: MIT
License Family: permissive
Topics: deep-learning, platform, pytorch
Last push: 2025-07-29T13:44:13+00:00

## Health v2 (maintenance only)
Score: 45/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 34, release rhythm 35, longevity 89
- inputs: {"age_days": 1252, "days_push": 400, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 25208, forks 2206 (observed 2026-08-28T04:11:38.058781+00:00)

## What it is
JARVIS (HuggingGPT) is a system that uses an LLM as a controller to plan tasks and orchestrate numerous expert ML models from the HuggingFace Hub as executors. It connects language models with the broader ML community to solve complex AI tasks across modalities.

## Use cases
- orchestrate multiple AI models with an LLM controller
- solve multimodal AI tasks via natural language
- build an LLM agent that calls HuggingFace models
- automate task planning and model selection with GPT
- run a server that routes requests to expert models
- benchmark LLM task automation capabilities

## When to choose
- you want an LLM to coordinate many HuggingFace expert models
- you need task planning, model selection, and execution in one pipeline
- you want to experiment with LLM-driven agent orchestration research

## When to avoid
- you need a lightweight production agent framework with minimal dependencies
- you don't have access to an OpenAI/Azure LLM endpoint or local model endpoints
- you need guaranteed low-latency responses, since multi-stage orchestration is slow

## Facets
- artifact type: framework
- maturity: active
- function: agent-framework, llm-inference, machine-learning, rag
- domain: artificial-intelligence, large-language-models, deep-learning
- platform: python, cli
- tags: hugginggpt, llm-orchestration, huggingface-models, task-planning, tool-use, pytorch, ai-agents, docker, web-server

## Member repositories
- microsoft/JARVIS (main) score 45

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:38.058781+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T16:56:05.254925+00:00, confidence not recorded.
  - readme: https://github.com/microsoft/JARVIS (fetched 2026-08-28T04:11:38.058781+00:00, sha 85131fa8d57a)
- Data as of 2026-08-30T08:39:29.467469+00:00.