# ucbrise/clipper

A low-latency prediction-serving system

Repository: https://github.com/ucbrise/clipper
Canonical: https://ross.abutalabs.com/products/clipper
Homepage: http://clipper.ai
Language: C++
License: Apache-2.0
License Family: permissive
Last push: 2021-04-26T17:27:01+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 3597, "days_push": 1955, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1420, forks 281 (observed 2026-08-28T04:04:40.584896+00:00)

## What it is
Clipper is a low-latency prediction serving system that sits between user-facing applications and machine learning models, exposing a standard REST interface for predictions and feedback. It supports multiple ML frameworks with adaptive batching, caching, and straggler mitigation, but is no longer actively maintained and is available only as a research artifact.

## Use cases
- serve machine learning model predictions over a REST API
- deploy models from different ML frameworks behind one interface
- reduce prediction latency with adaptive batching and caching
- collect prediction feedback from applications for model improvement
- combine predictions from multiple models with ensemble methods
- run a self-hosted model serving cluster with Docker

## When to choose
- you need a research-grade prediction serving system to study or extend
- you want a framework-agnostic REST layer in front of multiple ML models
- you need millisecond-latency serving with batching and caching techniques

## When to avoid
- you need a maintained, production-supported model server
- you want modern features like GPU inference servers or Kubernetes-native serving
- your stack requires recent Python versions or active community support

## Facets
- artifact type: service
- maturity: abandoned
- function: machine-learning, http-server, caching, monitoring, deployment
- domain: machine-learning, artificial-intelligence, backend, apis, microservices
- platform: python, cross-platform, self-hosted
- tags: prediction-serving, model-serving, model-deployment, low-latency, rest-api, research-artifact, adaptive-batching, ensemble-methods, docker

## Member repositories
- ucbrise/clipper (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:40.584896+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 2026-08-30T04:37:50.335045+00:00, confidence not recorded.
  - readme: https://github.com/ucbrise/clipper (fetched 2026-08-28T04:04:40.584896+00:00, sha c11c2c8476cb)
- Data as of 2026-08-30T08:39:29.467469+00:00.