This Half-Gigabyte AI Model Runs Local Agents on Your Phone

In brief MiniCPM5-1B scores an average of 42.57 across agentic and reasoning benchmarks, beating […]