IBM Research·released 2026-04-061 source
Not yet benchmarked
Claims drawn from cited facts, not live model generation.
This model was developed by IBM Research and released in 2026 with 28.87 billion parameters, placing it in the mid-size range rather than at the compact or frontier-scale extremes. That mid-size scale is a deliberate trade-off: it gives up some of the efficiency and lower cost of smaller models in exchange for more headroom and capability, along with the greater expense that comes with it.
4 cited facts
This model is not yet benchmarked, as it is scored on zero tracked benchmarks.
1 cited fact
The average gap to the state of the art is not disclosed, so no points-behind number can be stated and no frontier, competitive, or well-behind band can be assigned. It holds the top score on zero benchmarks, which means there is no evidence of category leadership on any of its tracked benchmarks yet.
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
granite-4.1-30b is an AI model developed by IBM Research, released 2026-04-06. Its benchmark record below tags every score as independently reproduced, self-reported, or unverified.