llm-speed: signed LLM inference-speed benchmarks
收藏官方服务：
资源简介：
Crowdsourced, cryptographically signed measurements of how fast large language models actually run: decode tokens per second, time to first token, and latency, across consumer GPUs, Apple Silicon, and hosted APIs, under one reproducible workload suite (suite-v1). Every run is signed on upload (EdDSA/JWS) so numbers cannot be edited after the fact. Live data and per-run permalinks: https://llm-speed.com/data. Methodology: https://llm-speed.com/methodology.
提供机构：
Zenodo创建时间：
2026-07-08
