Grok 4.6

NewGrok 4.6

Ranks #15 of 43 on reading real contracts into structured billing data — -4 pts vs GPT-5.5.

65.3%

Accuracy

#15 of 43

11.8%

Hallucination

HIGH-confidence & wrong

$108

Cost / 1k contracts

$0.11 each · via OpenRouter

140.2s

Median latency

p90 194.6s

±3

Run-to-run σ

3 runs

0.6

Value

acc. pts per $/1k

Tokens per contract

Input21,246

Output10,957

Reasoning7,151

Reliability 100% · valid structured output across 18 calls.

Want the full picture? How we score →