Gemini 3.6 Flash

NewGemini 3

Ranks #1 of 43 on reading real contracts into structured billing data — +9.6 pts vs GPT-5.5.

78.9%

Accuracy

#1 of 43

14.5%

Hallucination

HIGH-confidence & wrong

$116

Cost / 1k contracts

$0.12 each · via OpenRouter

52.8s

Median latency

p90 69.1s

±5.6

Run-to-run σ

3 runs

0.68

Value

acc. pts per $/1k

Tokens per contract

Input24,624

Output10,478

Reasoning6,931

Reliability 100% · valid structured output across 18 calls.

Want the full picture? How we score →