Artificial Analysis Launches Endpoint Accuracy Index v1.1

Artificial Analysis has launched Endpoint Accuracy Index v1.1, featuring expanded coverage for GLM-5.3 Flash and Muse Spark Open. The benchmark evaluates nearly 100 serverless endpoints as a percentage of a self-hosted reference deployment across tool calling, scientific reasoning, and long context recall. The v1.1 upgrade updates answer keys, increases test repeats to 25, and transitions grading to GPT-5.6 Luna to improve confidence intervals. Open article →