Published task benchmark

AML Narrative Writer

Convert Suspicious Activity Report (SAR) data into a narrative report.

Compare measured quality, workload cost, and latency across model tiers. Results are task-specific and directional—not a general model ranking.

Legal & ComplianceGenerationEvaluated 9/10/2026100 evaluation examplesFallback pricing checked September 11, 2026
Generation

AML Narrative Writer

Convert Suspicious Activity Report (SAR) data into a narrative report.

Legal & CompliancePer case
Not benchmarked
No result on this tier
Token profile pending
/100K
Run this tier to add it
Gemma 4 E4B IT
E4B · Off-the-shelf
1,448 in · 685 out · 11.14s p50
$10.28 /100K
Current price · TrustedRouter
View token rates · observed 9/11/2026
TrustedRouter · $0.0211 input · $0.1055 output /1M
81%
Cheapest
Mistral Large 2407
123B · Off-the-shelf
1,574 in · 879 out · 21.92s p50
$222 /100K
Best current price · TrustedRouter
Compare 2 router rates · observed 9/11/2026
TrustedRouter · $0.5275 input · $1.58 output /1M
OpenRouter · $2.00 input · $6.00 output /1M
89%
GPT-5.4
Hosted · Off-the-shelf
1,236 in · 833 out · 8.43s p50
$1559 /100K
Best current price · OpenRouter
Compare 2 router rates · observed 9/11/2026
OpenRouter · $2.50 input · $15.00 output /1M
TrustedRouter · $2.64 input · $15.82 output /1M
95%
Best accuracy
All 12 model results and methodology

Generated 100/100 examples

ModelTierQualityJudgedScenario cost
Nemotron Nano 9B v2small14%100/100$45.86
Qwen3 235B A22Bmid63%100/100$235
DeepSeek V3mid76%100/100$101
Mistral Large 2407mid89%100/100$222
Arcee Trinity Large Thinkingmid41%100/100$152
GPT-5.4frontier95%100/100$1559
Claude Opus 4.7frontier39%100/100$5852
Gemini 3.1 Pro Previewfrontier0%100/100$2682
Gemma 4 E4B ITsmall81%100/100$10.28
Granite 4.1 8Bsmall31%100/100$18.14
Ministral 8B Instruct 2410small33%100/100$39.38
Qwen3 4B Instruct 2507small5%100/100$122

LLM-judge pass rate on 100 synthetic examples. Generator: gpt-5.2. Judge: gpt-5.2. Evaluated 2026-09-10T17:34:02.158Z.

Directional: measured on a synthetic eval set generated by drydock. Cost/latency are not yet captured for taskrouter-run benchmarks.

drydock

Have a neighboring workload?

Describe it in Task Explorer. If it is not measured yet, you can request a benchmark and follow future results.

Explore another task