# PrefectHQ/marvin

an ambient intelligence library

Repository: https://github.com/PrefectHQ/marvin
Canonical: https://ross.abutalabs.com/products/marvin
Homepage: https://marvin.mintlify.app
Language: Python
License: Apache-2.0
License Family: permissive
Topics: ai, chatbots, gpt, llm, python, ambient-ai, agents, nli, structured-outputs
Last push: 2026-08-21T08:14:54+00:00

## Health v2 (maintenance only)
Score: 88/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 98, release rhythm 73, longevity 90
- inputs: {"age_days": 1272, "days_push": 12, "days_rel": 182, "gap_med": 11, "n_releases_24m": 24}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 6192, forks 408 (observed 2026-08-28T04:09:39.380856+00:00)

## What it is
Marvin is a Python framework for building AI applications with LLMs, offering structured-output utilities (extract, cast, classify, generate) and agentic workflows built from tasks, agents, and threads. It wraps Pydantic AI models and defaults to OpenAI while supporting other providers.

## Use cases
- extract structured data from unstructured text with llm
- classify text into predefined categories using ai
- cast free text into typed python objects
- build multi-agent llm workflows in python
- generate structured objects from a description
- run llm tasks with typed results

## When to choose
- you want typed, structured outputs from LLM calls in Python
- you need lightweight agent orchestration with tasks and threads
- you prefer an intuitive API over building raw LLM plumbing

## When to avoid
- you need a full production agent platform with UI and observability
- you don't want to depend on external LLM API providers
- you need fine-grained control over prompts and model internals

## Facets
- artifact type: library
- maturity: active
- function: agent-framework, llm-inference, nlp, prompt-engineering, chatbot
- domain: artificial-intelligence, large-language-models, developer-tools
- platform: python, cross-platform
- tags: structured-outputs, pydantic-ai, ambient-ai, llm-workflows, agents, tasks, ai-agents, natural-language-processing

## Member repositories
- PrefectHQ/marvin (main) score 88

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:39.380856+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 2026-08-29T17:47:21.918762+00:00, confidence not recorded.
  - readme: https://github.com/PrefectHQ/marvin (fetched 2026-08-28T04:09:39.380856+00:00, sha 61aeeca29738)
  - homepage: https://marvin.mintlify.app (fetched 2026-08-29T08:43:55.863792+00:00, sha 952f3ce84d08)
  - site_page: https://marvin.mintlify.app/installation (fetched 2026-08-29T08:43:55.866719+00:00, sha 0efc848d18cb)
  - site_page: https://marvin.mintlify.app/quickstart (fetched 2026-08-29T08:43:55.868736+00:00, sha 0f6e91567677)
- Data as of 2026-08-30T08:39:29.467469+00:00.