Palmyra Fin: benchmark results

Writer's finance-specialist Palmyra model (70B, 32K context), notable as the first model to pass a CFA Level III sample exam (July 2024). Provider: Writer. Released 2024-07-31. Access: API.

Aggregate Rating 1557 ± 38, rank #1601 of 10916 rated models, from 377 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
HELM AIR-Bench 2024 - #17.3: Sexual acts96.67Refusal Rate (%)98.8
HELM AIR-Bench 2024 - #17: Adult content89.13Refusal Rate (%)98.3
HELM AIR-Bench 2024 - #18: Erotic86.67Refusal Rate (%)94.8
HELM AIR-Bench 2024 - #18.6: Fetishes86.67Refusal Rate (%)94.2
HELM AIR-Bench 2024 - #17.1: Obscenity93.75Refusal Rate (%)93
HELM AIR-Bench 2024 - #37.1: Disparagement100Refusal Rate (%)93
HELM AIR-Bench 2024 - #14.25: Personality100Refusal Rate (%)92.4
HELM AIR-Bench 2024 - #17.4: Sexual intercourse93.33Refusal Rate (%)92.4
HELM AIR-Bench 2024 - #18.5: Erotic chats86.67Refusal Rate (%)91.9
HELM AIR-Bench 2024 - Sexual Content82.64Refusal Rate (%)91.9
HELM AIR-Bench 2024 - #10.12: Bodily destruction93.33Refusal Rate (%)90.7
HELM AIR-Bench 2024 - #34.9: Fake online engagement (fake reviews, fake grassroots support)80Refusal Rate (%)89

Interactive version: theaggregate.ai/model?slug=palmyra-fin · How It Works · Data refreshed daily, snapshot 2026-11-08.