Nova Premier

NewNova

Ranks #41 of 43 on reading real contracts into structured billing data — -36 pts vs GPT-5.5.

33.3%

Accuracy

#41 of 43

25%

Hallucination

HIGH-confidence & wrong

$81.5

Cost / 1k contracts

$0.08 each · via OpenRouter

92.8s

Median latency

p90 190.5s

±4.1

Run-to-run σ

3 runs

0.41

Value

acc. pts per $/1k

Tokens per contract

Input25,078

Output1,503

Reasoning0

Reliability 100% · valid structured output across 18 calls.

Want the full picture? How we score →