# rockchip-linux/rknn-toolkit2

Repository: https://github.com/rockchip-linux/rknn-toolkit2
Canonical: https://ross.abutalabs.com/products/rockchip-linux-rknn-toolkit2
Language: C
License: BSD-3-Clause
License Family: permissive
Last push: 2024-03-25T08:23:39+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 1919, "days_push": 891, "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 1193, forks 179 (observed 2026-08-28T04:03:56.447380+00:00)

## What it is
RKNN-Toolkit2 is Rockchip's software development kit for converting trained AI models to RKNN format and deploying them on Rockchip NPU chips like RK3588 and RV1106. It includes PC-side model conversion and performance evaluation tools, Python deployment APIs, and a C/C++ runtime for on-device inference.

## Use cases
- convert onnx models to run on rockchip npu
- deploy deep learning models on rk3588
- run inference on embedded ai hardware
- benchmark model performance on npu
- deploy computer vision models on rv1106

## When to choose
- you are deploying AI models to Rockchip RK3566/RK3568/RK3588/RK3562 or RV1103/RV1106 chips
- you need to convert ONNX models to an NPU-optimized format
- you want C/C++ or Python inference APIs on Rockchip hardware

## When to avoid
- you target older Rockchip chips like RK1808, RV1109, RV1126, or RK3399Pro (use rknn-toolkit instead)
- you need a maintained repository - this one has moved to airockchip/rknn-toolkit2
- you deploy to non-Rockchip hardware such as GPUs or other NPUs

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, llm-inference, sdk, compiler
- domain: machine-learning, embedded-systems, deep-learning, hardware
- platform: python, cpp, embedded
- tags: npu, rockchip, model-conversion, edge-ai, rknn, inference, linux

## Member repositories
- rockchip-linux/rknn-toolkit2 (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:56.447380+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 2026-08-30T06:22:26.191958+00:00, confidence not recorded.
  - readme: https://github.com/rockchip-linux/rknn-toolkit2 (fetched 2026-08-28T04:03:56.447380+00:00, sha fc70c169a59a)
- Data as of 2026-08-30T08:39:29.467469+00:00.