Analysis

Match invoice, PO, and receipt records; calculate variances; approve or route exceptions.

Finance, accounting, and procurement operationsPer invoice

Not benchmarked

No result on this tier

Token profile pending

— /100K

Run this tier to add it

—

Nemotron Nano 9B v2

9B · Off-the-shelf

753 in · 1,466 out · 9.39s p50

$38.24 /100K

AWS Bedrock temporary list price

View token rates

AWS Bedrock · $0.0600 input · $0.2300 output /1M

74%

Qwen3 235B A22B

235B (22B active) · Off-the-shelf

714 in · 1,082 out · 15.34s p50

$229 /100K

Best current price · OpenRouter

Compare 2 router rates · observed 9/11/2026

OpenRouter · $0.4550 input · $1.82 output /1M

TrustedRouter · $0.6330 input · $2.11 output /1M

76%

Claude Opus 4.7

Hosted · Off-the-shelf

1,060 in · 1,384 out · 14.70s p50

$3990 /100K

Best current price · OpenRouter

Compare 2 router rates · observed 9/11/2026

OpenRouter · $5.00 input · $25.00 output /1M

TrustedRouter · $5.28 input · $26.38 output /1M

94%

Best accuracy

All 12 model results and methodology

Generated 100/100 examples

Model	Tier	Quality	Judged	Scenario cost
Nemotron Nano 9B v2	small	74%	100/100	$38.24
Qwen3 235B A22B	mid	76%	100/100	$229
DeepSeek V3	mid	73%	100/100	$82.85
Mistral Large 2407	mid	75%	100/100	$193
Arcee Trinity Large Thinking	mid	48%	100/100	$143
GPT-5.4	frontier	92%	100/100	$1461
Claude Opus 4.7	frontier	94%	100/100	$3990
Gemini 3.1 Pro Preview	frontier	91%	100/100	$1876
Gemma 4 E4B IT	small	68%	100/100	$9.95
Granite 4.1 8B	small	55%	100/100	$12.23
Ministral 8B Instruct 2410	small	24%	100/100	$26.39
Qwen3 4B Instruct 2507	small	63%	100/100	$99.22

LLM-judge pass rate on 100 synthetic examples. Generator: gpt-5.2. Judge: gpt-5.2. Evaluated 2026-09-10T23:27:02.489Z.

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?