- Parameters
- 8B
- Quant
- Q4_K_M GGUF, and MLX mixed_3_6 at 3.8GB
- Size
- 5.1 GB
- Family
- granite
granite-guardian-4.1-8b
Seated · Screendenseresidentllama.cppmlxomlx
Model card ↗Outcome as of 20 September 2026, from the seat registry. The write-up below was made when the model was tested.
Why it was tested
A purpose-built safety and screening classifier, tested for the judge seat where a general model was overkill.
What came of it
100% on screening at 5.1GB. Worth noting the caveat: it is a specialist scored on a suite close to saturation, so the perfect result says less than it appears to. Converted to MLX mixed_3_6 on 2026-09-09 at 3.8GB, chosen over plain 4-bit because uniform quantisation destroyed tool calling on REAP (5/12 against 12/12) and a judge's whole job is precision. That conversion is not yet validated against the 100% baseline.
| Suite | Score | 95% interval | Passed | Runtime · profile | Recorded |
|---|---|---|---|---|---|
| screening | 100.0% | 94.0–100.0 | Passed 60/60 | llama.cppextraction | 2026-08-31 |
Runs marked unenforced were measured through mlx_lm on a suite that requires constrained decoding. mlx_lm accepts the schema and ignores it, so those numbers do not measure what the suite is designed to measure. They are kept and labelled rather than deleted.
Every run log entry for this model (1)
| Finished | Suite | Runtime · profile | Score | Coverage | Grader · dataset · config |
|---|---|---|---|---|---|
| screeningscreening/1 | llama.cppextraction | 100.0% | Coverage 60/60 | g 0661b49a10be2d2ad 64c97ca2269c29afc 9ba7cd65a17e6418 |
- Granite 4.2 30BMeasured, not seated
30B · dense · Q4_K_M · 16.5 GB
Tested because IBM builds explicitly for structured JSON output, and nullschema is the weakest axis across the whole fleet.
- 2026-09-14Q4_0 decodes up to 2x faster than Q4_K_M at the same bit-width, and emits valid JSON more reliablyrevised
- 2026-09-14Byte alignment, not bit count, governs decode speedrevised
- 2026-09-09Judge and tool-use scores are runtime-sensitive because the decisions sit on ~0.4 logprob margins
- 2026-09-09oMLX degrades models it has no native kernel for, by up to 58 points
- 2026-09-09Converting the judge to MLX destroyed it, from 100% to 36.7%