OpenBMB Releases MiniCPM5-2B: A 2.52B Dense Model Averaging 53.9 Across 34 Benchmarks and Built to Run On Device

OpenBMB has released MiniCPM5-2B, a dense language model containing 2.52 billion parameters and a native context window of 131,072 tokens. Designed for on-device performance, the model achieves a 53.9 average across 34 benchmarks, notably outperforming larger models like Qwen3.5-4B in coding, tool use, and long-context retrieval tasks.