openbmb·released 2026-05-211 source
Not yet benchmarked
Claims drawn from cited facts, not live model generation.
This model comes from OpenBMB and, at 1.08 billion parameters, is a compact model released in 2026. That compact scale trades raw capacity for efficiency, keeping inference and deployment costs lower while leaving less headroom than large or frontier-scale alternatives.
4 cited facts
This model is not yet benchmarked: it is scored on no tracked benchmarks. With no benchmark scores on record, it holds no current top score.
2 cited facts
No average SOTA gap is available for this model, so its standing cannot be expressed as a point gap under the disclosed harness; what is known is that it holds the top score on zero benchmarks. That plain reading means it has no category leadership yet on any of the benchmarks counted — not a frontier claim, just the absence of a first-place result.
3 cited facts
2 facts cross-checked across data sources: 2 single-source.
Source facts, citations, and refresh stamp for this record.
Sources: huggingface_models
MiniCPM5-1B is an AI model developed by openbmb, released 2026-05-21. Its benchmark record below tags every score as independently reproduced, self-reported, or unverified.