MLPerf Client
mlcommons.org
Spec sheet
- Purpose
- MLPerf Client benchmarks LLM, generative image, and agentic AI performance on personal computers, including laptops, desktops, and workstations.Read 9 Oct 2026 · mlcommons.org
- Free plan
- YesRead 23 Sep 2026 · mlcommons.org
- Support
- Users can submit questions, feature requests, or problems through the support form for working-group review and response as needed.Read 9 Oct 2026 · mlcommons.org
- Maker
- MLCommons describes itself as an open AI engineering consortium that develops benchmarks through collaboration between industry and academia.Read 9 Oct 2026 · mlcommons.org
- Agentic AI
- Agentic tests cover Software Engineering and Data Analyst tasks and measure end-to-end duration, including LLM inference and tool execution.Read 9 Oct 2026 · mlcommons.org
- CLI or automation
- YesRead 23 Sep 2026 · mlcommons.org
- Custom test profiles
- YesRead 23 Sep 2026 · mlcommons.org
- Download and source
- The benchmark is downloadable from its GitHub release page, and its public GitHub repository contains source code that users may inspect or modify.Read 8 Oct 2026 · mlcommons.org
- Downloads
- The benchmark is downloadable from GitHub, its iPad version is available in Apple's App Store, and its public GitHub repository contains source code.Read 9 Oct 2026 · mlcommons.org
- Experimental components
- The Linux CLI version, power-efficiency tools, Flux 2 Klein 4B image generation, and some agentic workloads are identified as experimental.Read 8 Oct 2026 · mlcommons.org
- Hardware limits
- The benchmark requires 200GB of free drive space and recommends 400GB to run all models; minimum hardware and memory requirements vary by platform.Read 8 Oct 2026 · mlcommons.org
- Hardware requirements
- Minimum hardware requirements list selected AMD, Intel, NVIDIA, Qualcomm, and Apple systems, with memory and VRAM requirements varying by platform.Read 9 Oct 2026 · mlcommons.org
- Image generation
- The image generation test uses the experimental Flux 2 Klein 4B model and reports images per minute at 1024×1024 resolution.Read 9 Oct 2026 · mlcommons.org
- Internet use
- An internet connection is needed for the first run with an acceleration path; downloaded files are stored locally for later runs.Read 8 Oct 2026 · mlcommons.org
- LLM workloads
- Its LLM tasks include creative writing, content generation, structured text, code analysis, and summarization, using models including Llama 3.1 8B Instruct and Phi 4 Mini Instruct.Read 9 Oct 2026 · mlcommons.org
- Maker history
- MLCommons' history page lists its launch as December 3, 2020.Read 9 Oct 2026 · mlcommons.org
- Network use
- An internet connection is needed for the first run with an acceleration path to download files; later runs use locally stored files.Read 9 Oct 2026 · mlcommons.org
- Operating systems
- Windows 11; macOS Tahoe 26; iPadOS 26; Ubuntu Linux 24.04Read 23 Sep 2026 · mlcommons.org
- Quality checks
- The working group applies MMLU and IFEval accuracy requirements to qualify accelerated execution paths and models for release.Read 9 Oct 2026 · mlcommons.org
- Result export formats
- CSVRead 23 Sep 2026 · mlcommons.org
- Result validity
- Only scores produced with configurations marked “Configuration tested by MLCommons” are considered valid MLPerf Client scores.Read 8 Oct 2026 · mlcommons.org
- Scoring limits
- Only base components can contribute to an official MLCommons-tested score; extended and experimental results are marked separately.Read 9 Oct 2026 · mlcommons.org
- Storage requirement
- The benchmark requires 200GB of free drive space and recommends 400GB to run all models.Read 9 Oct 2026 · mlcommons.org
- Supported systems
- The stated operating system requirements are Windows 11, macOS Tahoe 26, iPadOS 26, or Ubuntu Linux 24.04; Linux is described as CLI-only and experimental.Read 9 Oct 2026 · mlcommons.org
- Workload areas
- LLM inference; image generation; agentic AIRead 23 Sep 2026 · mlcommons.org
Where Specifiction places it
- PC benchmark tools#39 of 707.7
- PC Benchmark Software#47 of 737.6