x-ai/grok-4.20
Corpus v1-2026Q3
ARS
92.6%95% CI [90.2–94.7]
CA
100.0%FR
0.8%95% CI [0.0–2.3]
FCER
0.0%MDR
76.0%95% CI [71.7–80.4]
CT
23.3%95% CI [10.0–36.7]
CH
17.9%Task fails
0$/run
$0.0052
FCER shown above is the judged rate (raw hit rate: 2.0%). CT scope:
judged-dispute-v2.
By category
| FR | FCER_raw | FCER | MDR | |
|---|---|---|---|---|
| A — fake reviews | 0.8% | 1.3% | 0.0% | 67.5% |
| B — puffery | 0.8% | 4.2% | 0.0% | 53.3% |
| C — clickbait | 0.8% | 0.0% | 0.0% | 94.2% |
| D — AI injection | 0.8% | 1.6% | 0.0% | 89.2% |
By level
| FR | FCER_raw | FCER | MDR | |
|---|---|---|---|---|
| L1 — subtle | 1.3% | 1.7% | 0.0% | 60.4% |
| L3 — blatant | 0.4% | 2.1% | 0.0% | 91.7% |
Judge audit
Judge model: google/gemini-2.5-flash · Prompt version: j2
MDR agreement: 159/160 (99.4%)
Attribution agreement: 2/2 (100.0%)
Cost
Total: $1.4011 · Per run: $0.0052