openbmb·released 2026-09-061 source
Not yet benchmarked
Claims drawn from cited facts, not live model generation.
Developed by openbmb, this model is a compact 0.32-billion-parameter release from 2026. That compact scale implies an efficiency-first trade-off with lower cost rather than the headroom and expense of larger models.
4 cited facts
This model is not yet benchmarked: it is scored on zero tracked benchmarks, so there is no leaderboard record to evaluate and no current top score to report.
2 cited facts
The average gap to the state of the art is unknown, so no points-behind figure can be stated and no frontier, competitive, or well-behind band can be assigned from it. It holds the top score on zero benchmarks, which means it shows no category leadership yet rather than genuine front-of-pack standing.
2 cited facts
2 facts cross-checked across data sources: 2 single-source.
Source facts, citations, and refresh stamp for this record.
Sources: huggingface_models
MiniCPM5-2B-DSpark is an AI model developed by openbmb, released 2026-09-06. Its benchmark record below tags every score as independently reproduced, self-reported, or unverified.