# superduper-io/superduper

Superduper: End-to-end framework for building custom AI applications and agents.

Repository: https://github.com/superduper-io/superduper
Canonical: https://ross.abutalabs.com/products/superduper
Homepage: https://superduper.io
Language: Python
License: Apache-2.0
License Family: permissive
Topics: ai, mlops, torch, transformers, mongodb, python, pytorch, ml, database, data, inference, distributed-ml, llm-inference, pretrained-models, chatbot, semantic-search, llm-serving, llmops, vector-search, rag
Last push: 2025-09-01T15:20:18+00:00

## Health v2 (maintenance only)
Score: 49/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 39, release rhythm 32, longevity 100
- inputs: {"age_days": 1464, "days_push": 366, "days_rel": 372, "gap_med": 0.5, "n_releases_24m": 63}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 5316, forks 544 (observed 2026-08-28T04:09:15.536723+00:00)

## What it is
Superduper is an open-source Python framework for building database-integrated AI applications and agents, connecting models, embeddings, and vector search directly to existing databases like MongoDB, SQL, Snowflake, and Redis. It supports end-to-end workflows including RAG, semantic search, model inference, and agent orchestration with self-hosted or cloud deployment options.

## Use cases
- build a RAG pipeline over my existing database
- add semantic search to my MongoDB data
- generate vector embeddings for documents in SQL tables
- deploy AI agents that query my enterprise databases
- run LLM inference integrated with my data stack
- build a chatbot that answers questions from company documents
- streamline MLOps for deploying pretrained models on my data

## When to choose
- you want AI features (search, RAG, agents) directly on top of existing databases without moving data
- you need a Python-native framework supporting PyTorch, Transformers, and multiple database backends
- you want self-hosted or on-prem AI deployment with data security requirements
- you're building enterprise AI agents across structured and unstructured data

## When to avoid
- you need a simple one-off vector store without database integration
- you're looking for a managed SaaS-only solution with no self-hosting
- your stack doesn't include any supported database backend
- you need a lightweight inference server rather than a full application framework

## Facets
- artifact type: framework
- maturity: active
- function: machine-learning, llm-inference, rag, vector-database, agent-framework, database, search-engine, chatbot, etl
- domain: artificial-intelligence, machine-learning, large-language-models, databases, developer-tools
- platform: python, self-hosted, cloud
- tags: vector-search, semantic-search, mlops, llmops, embeddings, pytorch, transformers, mongodb, snowflake, in-database-ai, retrieval-augmented-generation, ai-agents, data-engineering, docker, kubernetes

## Member repositories
- superduper-io/superduper (main) score 49

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:15.536723+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 2026-08-29T17:58:57.165065+00:00, confidence not recorded.
  - readme: https://github.com/superduper-io/superduper (fetched 2026-08-28T04:09:15.536723+00:00, sha 5fa4ad967982)
  - homepage: https://superduper.io (fetched 2026-08-29T08:53:44.229045+00:00, sha 29c47743bcf0)
  - site_page: https://docs.superduper.io/ (fetched 2026-08-29T08:53:44.240451+00:00, sha a797a4eb837d)
  - site_page: https://superduper.io/about (fetched 2026-08-29T08:53:44.242226+00:00, sha 06e585de739c)
  - site_page: https://docs.superduper.io (fetched 2026-08-29T08:53:44.244242+00:00, sha a797a4eb837d)
  - site_page: https://superduper.io/pricing (fetched 2026-08-29T08:53:44.238554+00:00, sha 059de9b4a0b9)
- Data as of 2026-08-30T08:39:29.467469+00:00.