Primate Labs has launched Geekbench AI 1.0, a new benchmarking platform designed to assess the artificial intelligence (AI) performance of various devices. This free app is now available for download across all major platforms. Geekbench AI 1.0 evaluates a device’s AI capabilities by running a series of tests on the CPU, GPU, and neural processing unit (NPU) to generate a comprehensive performance score. Developers can also select specific AI frameworks and models to test different workloads.
The company announced the launch through a blog post, describing Geekbench AI as a benchmarking suite focused on machine learning, deep learning, and AI-driven tasks. The platform is known for its consistent cross-platform utility and real-world workload reflection.
The app automatically runs ten distinct AI workloads, each requiring three different data types, allowing users to obtain a thorough assessment of a device’s AI performance. It supports Android, iOS, Linux, macOS, and Windows, making it compatible with smartphones, tablets, laptops, desktops, and other similar devices.
Initially introduced as Geekbench ML, the app was renamed to reflect the industry’s shift toward using “AI” to describe these workloads. The benchmarking tool considers various factors such as workloads, hardware, and AI frameworks to accurately measure AI performance.
Geekbench AI also evaluates the speed and accuracy of a device, helping to identify any trade-offs between performance and efficiency. It measures various metrics, including datasets, frameworks, runtime, computer vision, and natural language processing (NLP).
Additionally, Primate Labs has introduced an ML Benchmarks leaderboard, where users can compare the CPU, GPU, and NPU performance of different devices and identify top performers. The app has specific minimum software requirements:
Android: Android 12 or later, 4GB RAM
iOS: iOS 17 or later
Linux: Ubuntu 22.04 LTS (64-bit) or later, 4GB RAM (AMD or Intel processor)
macOS: macOS 14 or later, 8GB RAM (Apple Silicon or Intel processor)
Windows: Windows 10 (64-bit) or later, 8GB RAM (AMD, ARM, or Intel processor)