llmspeed

llmspeed lets you select hardware and language models to see simulated token generation speeds and time-to-first-token metrics. It uses verified benchmarks for GPUs, Macs, and other hardware to help you understand real-world performance of local LLM deployments before making hardware purchases.