# punica-ai/punica

Serving multiple LoRA finetuned LLM as one

Repository: https://github.com/punica-ai/punica
Canonical: https://ross.abutalabs.com/products/punica
Homepage: https://arxiv.org/abs/2310.18547
Language: Python
License: Apache-2.0
License Family: permissive
Topics: large-language-models, llm, lora
Last push: 2024-05-08T12:25:17+00:00

## Health v2 (maintenance only)
Score: 19/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 79
- inputs: {"age_days": 1113, "days_push": 847, "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 1175, forks 64 (observed 2026-08-28T04:03:52.496068+00:00)

## What it is
Punica is a Python system for serving many LoRA-finetuned LLMs from a single copy of the base model on one GPU, using a custom CUDA kernel (SGMV) to batch LoRA adapter computations. It achieves up to 12x higher throughput than state-of-the-art LLM serving systems for multi-tenant LoRA workloads.

## Use cases
- serve multiple LoRA finetuned LLMs on one GPU
- batch requests for different LoRA adapters efficiently
- reduce GPU memory when hosting many finetuned models
- multi-tenant LLM serving with per-user adapters
- benchmark LoRA serving throughput against vLLM or DeepSpeed

## When to choose
- you need to serve many LoRA variants of the same base model cost-effectively
- GPU memory is a bottleneck from loading many finetuned model copies
- your workload mixes requests across different LoRA adapters

## When to avoid
- you serve only a single model with no LoRA adapters
- you need a production system with active development and broad model support
- you have no CUDA-capable GPU

## Facets
- artifact type: library
- maturity: maintenance
- function: llm-inference, gpu-computing, machine-learning
- domain: large-language-models, machine-learning, gpu-computing
- platform: python
- tags: lora, multi-tenant-serving, cuda-kernels, sgmv, llm-serving, gpu, linux

## Member repositories
- punica-ai/punica (main) score 19

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:52.496068+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 2026-08-30T06:28:14.289227+00:00, confidence not recorded.
  - readme: https://github.com/punica-ai/punica (fetched 2026-08-28T04:03:52.496068+00:00, sha 120d2787e0d9)
  - homepage: https://arxiv.org/abs/2310.18547 (fetched 2026-08-29T12:33:32.425651+00:00, sha 4b09fba9121e)
  - site_page: https://info.arxiv.org/about/donate.html (fetched 2026-08-29T12:33:32.434612+00:00, sha cca9c3a11c56)
  - site_page: https://info.arxiv.org/about/ourmembers.html (fetched 2026-08-29T12:33:32.438372+00:00, sha 47cbc55ff1de)
  - site_page: https://info.arxiv.org/about (fetched 2026-08-29T12:33:32.441797+00:00, sha a1f16f915a9a)
  - site_page: https://info.arxiv.org/labs/index.html (fetched 2026-08-29T12:33:32.436446+00:00, sha b14a8d05a0ec)
- Data as of 2026-08-30T08:39:29.467469+00:00.