Unverified50% confidenceBenchmarkExact time
Among 51 test messages, all 46 messages with amounts recorded the correct amount and category, for a parsing accuracy of 100% with zero failures.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/14/2026
First Seen
Valid until: 1/12/2027
Sources
Related Entities
Related Claims
UnverifiedAll four models correctly read all 48 clean invoices and 48 placebo invoices, with zero errors across 4,704 scoring responses, and 0 of the 521 fabricated numbers happened to equal the true total that was obscured.72% similarUnverifiedOf the 962 responses, 51 failed, distributed as 26 truncations, 13 boundary errors, 12 paraphrases, and 0 corrections.70% similarUnverifiedThe accuracy data in the test has an error margin of approximately ±8–10 percentage points at n=100 sample size (based on a 95% confidence-level binomial confidence interval).69% similarUnverifiedFrom a response-conditioned diagnostic perspective, 58 of the 60 requests returned a response, and all 58 of those responses were rated as correct (clean 29/29, injected 29/29).66% similarUnverified41个完整测试套件中有20个拿到满分100%，该基准已无法区分这些领先模型65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1024720API
curl https://kongchang.com/api/v1/knowledge/claims/1024720MCP
get_claim(id=1024720)