# NVIDIA/kvpress

LLM KV cache compression made easy

Repository: https://github.com/NVIDIA/kvpress
Canonical: https://ross.abutalabs.com/products/kvpress
Language: Python
License: Apache-2.0
License Family: permissive
Topics: llm, inference, kv-cache, kv-cache-compression, long-context, python, pytorch, transformers, large-language-models
Last push: 2026-09-02T07:41:41+00:00

## Health v2 (maintenance only)
Score: 86/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 100, release rhythm 91, longevity 47
- inputs: {"age_days": 665, "days_push": 0, "days_rel": 62, "gap_med": 13.5, "n_releases_24m": 27}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1201, forks 174 (observed 2026-09-03T02:15:06.827285+00:00)

## What it is
kvpress is a Python library from NVIDIA that implements multiple KV cache compression methods and benchmarks for long-context LLM inference, built on Hugging Face transformers. It provides 'presses' that compress the key-value cache during prefilling to reduce memory usage, and simplifies developing new compression methods.

## Use cases
- compress kv cache for long-context llm inference
- reduce gpu memory when serving llms with long prompts
- benchmark kv cache compression methods
- implement a custom kv cache compression method
- run 1M token contexts on limited gpu memory
- integrate kv cache compression into a transformers pipeline

## When to choose
- you deploy or fine-tune transformer LLMs with long contexts and hit KV cache memory limits
- you research KV cache compression and want a benchmarking framework with existing methods
- you already use Hugging Face transformers and PyTorch

## When to avoid
- you need KV cache compression for non-transformer or non-Hugging Face model runtimes
- you want a production inference server rather than a compression library
- your contexts are short enough that KV cache memory is not a bottleneck

## Facets
- artifact type: library
- maturity: active
- function: llm-inference, machine-learning, benchmarking
- domain: large-language-models, deep-learning, machine-learning
- platform: python, cross-platform
- tags: kv-cache, kv-cache-compression, long-context, transformers, pytorch, huggingface, memory-optimization, gpu

## Member repositories
- NVIDIA/kvpress (main) score 86

## Provenance
- Observed fields: from GitHub, fetched 2026-09-03T02:15:06.827285+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 2026-08-30T06:22:12.262196+00:00, confidence not recorded.
  - readme: https://github.com/NVIDIA/kvpress (fetched 2026-09-03T02:15:06.827285+00:00, sha aeef658a5dcb)
- Data as of 2026-08-30T08:39:29.467469+00:00.