v3.8 Release

Accelerate YourAI PerformanceTesting

Configure unlimited combinations of models, software, hardware, and hyperparameters in seconds.

Start BenchmarkingSee Features
Now Supporting NVIDIA DGX™ Spark

Latest Features Available

Side-by-side benchmarking, smarter user support agent, support for benchmarking your own inference endpoint, and more.

Versus Workspace – Side-by-Side Benchmarking

Supports dual-system, side-by-side benchmarking for instant comparison with real-time charts for GPU/CPU usage, throughput, latency, and telemetry metrics.

User Support Agent Upgrades

User support agent now supports chat-based project creation, management, and execution, as well as auto-generation of run analysis, performance summaries, and debugging insights.

Bring Your Own Endpoint (BYOE)

Support for users to bring their own custom inference endpoints into Metrum Insights, and benchmark them on varying concurrency levels, different input prompt and output response lengths, and accuracy evaluation datasets.

Fully Automated AI Performance Benchmarking

Configure unlimited combinations of models, software, hardware, and hyperparameters in seconds.

Parameters

Configure and test across multiple dimensions simultaneously.

Concurrency Levels
Token Lengths
Request Rates
Precision Modes

Chips

Benchmark across architectures.

NVIDIA DatacenterAMD InstinctIntel Gaudi 3AMD EPYCIntel XeonRTX GPUsIntel Arc

Models

Test the latest foundation models.

GPT-OSSDeepSeekGemmaPhiQwenMistralLlama 4+

Real-Time Metrics

Track performance across every dimension.

Throughput

tok/s

Latency

TTFT/TPOT

Power Usage

Watts

Efficiency

tok/W

AI-Powered

AI-Powered Analysis with Creator

Leverage AI-powered analysis to automatically generate insights, identify bottlenecks, and receive optimization recommendations.

  • Automated report generation
  • Performance anomaly detection
  • Optimization suggestions
  • Natural language queries
Try Creator
AI-Powered Creator
Hardware Sizer

Hardware Planning

Smart Hardware Recommendations with Hardware Sizer

Get intelligent hardware recommendations based on your workload requirements.

GPU Selection
Cluster Sizing
Cost Analysis
Performance Estimates

Real-Time Monitoring

Performance Visualization with Pulse

Real-time visualization of performance metrics across your benchmarking runs.

Pulse Performance Visualization

Ready to Accelerate Your AI Performance?

Join industry leaders using Metrum Insights to optimize their AI infrastructure.

Get Started FreeView Documentation