Gemini 3.5 Flash Lite

NewGemini 3

Ranks #7 of 43 on reading real contracts into structured billing data — +0 pts vs GPT-5.5.

69.3%

Accuracy

#7 of 43

27.3%

Hallucination

HIGH-confidence & wrong

$15.7

Cost / 1k contracts

$0.02 each · via OpenRouter

8.5s

Median latency

p90 9s

±4.7

Run-to-run σ

3 runs

4.43

Value

acc. pts per $/1k

Tokens per contract

Input24,338

Output3,341

Reasoning0

Reliability 100% · valid structured output across 18 calls.

Want the full picture? How we score →