llm-smartrouter-benchmark

简介

LLM SmartRouter Latency & Cost Benchmark (2026) Empirical performance benchmark dataset comparing direct model endpoints (OpenAI, Anthropic, Google) against the PixelRouter / BLUN SmartRouter proxy layer (https://api.pixeloffice.eu/v1). Schema query_type: Task classification (code_generation, deep_reasoning, fast_chat, fact_retrieval). target_model: Routed model identifier. ttft_ms: Time-To-First-Token in milliseconds. total_latency_ms: End-to-end response… See the full description on the dataset page: https://huggingface.co/datasets/pixeloffice/llm-smartrouter-benchmark.