Command A

NewCommand

Ranks #30 of 43 on reading real contracts into structured billing data — -13.6 pts vs GPT-5.5.

55.7%

Accuracy

#30 of 43

31.6%

Hallucination

HIGH-confidence & wrong

$90.3

Cost / 1k contracts

$0.09 each · via OpenRouter

52.1s

Median latency

p90 61s

±6.5

Run-to-run σ

3 runs

0.62

Value

acc. pts per $/1k

Tokens per contract

Input24,315

Output2,949

Reasoning0

Reliability 100% · valid structured output across 18 calls.

Want the full picture? How we score →