Single-photon detectors have the highest sensitivity of light detection. The utilization of single-photon detectors in LiDAR system can greatly improve the comprehensive performance of the system. Laser in the second near-infrared region (1.0-1.7 μm) has the advantages of high atmospheric transmittance, weak scattering and weak solar background radiation, which is the ideal working band of aerosol remote sensing and three-dimensional imaging LiDAR system. Therefore, a high-performance miniaturized free-running single-photon detector is designed in this paper.The single-photon detector is based on InGaAs/InP negative feedback avalanche photodiode (NFAD), allowing it to operate in the free-running mode (Fig.1). A precise bias circuit and a precise temperature control circuit provide the bias voltage and cooling for the NFAD, respectively (Fig.2, Fig.3). In order to meet the needs of photon time-of-flight measurement for LiDAR system, the time-to-digital converter (TDC) function is realized by FPGA based on carry delay chain (Fig.4). Through the built-in micro controller unit (MCU) with integrated counting rate and afterpulse correction algorithm, it can make real-time correction of TDC data and output via USB interface.The detector has dimensions of 116 mm×107.5 mm×80 mm (Fig.5). The maximum detection efficiency is more than 35% at 1.5 μm (Fig.6), and the time jitter (full width at half maxima, FWHM) is as low as 80 ps (Fig.7). The time measurement accuracy of internal TDC can reach 100 ps. The miniaturized LiDAR product using this single-photon detector can detect up to 15 km with a range resolution of less than 30 m (Fig.8).The QCD600 series miniature free-running InGaAs/InP single-photon detector provides a compact and real-time data post-processing single-photon detection solution for LiDAR systems in the near infrared band with high efficiency, low noise, low time jitter. In the future, free-running single-photon detector will be developed in the direction of miniaturization using integrated refrigeration technology and ultra-low noise using deep refrigeration technology, which will provide more powerful technical support for LiDAR, QKD and other applications.