Will's Blog: SemiAnalysis launches InferenceMAX, an open-source benchmark that automatically tracks LLM inference performance across AI models and frameworks every night (SemiAnalysis)

10 October 2025

SemiAnalysis launches InferenceMAX, an open-source benchmark that automatically tracks LLM inference performance across AI models and frameworks every night (SemiAnalysis)

SemiAnalysis:
SemiAnalysis launches InferenceMAX, an open-source benchmark that automatically tracks LLM inference performance across AI models and frameworks every night — NVIDIA GB200 NVL72, AMD MI355X, Throughput Token per GPU, Latency Tok/s/user, Perf per Dollar, Cost per Million Tokens …

Posted from: this blog via Microsoft Power Automate.