BasedAGIBasedAGI
Menu
Rankings live

insurance

Fraud signal summary

Summarize potential fraud indicators with conservative evidence framing.

#1 Recommendation

gemini-2.5-pro

Strong on FACTS Benchmark Suite facts_grounding_score_pct (100%) and Vectara HHEM Leaderboard overall_hallucination_error_pct (76%)

external/google/gemini-2-5-pro

27.0%

Score

41.6%

Confidence

Limited benchmark evidence for this use case.

51 ranked models with average evidence of 15.3 points. Rankings may shift as more benchmark data is ingested.

Ranked Models

30

Evidence Quality

81%

Scoring

Benchmark-backed

Top Signal

FACTS Benchmark Suite: facts_grounding_score_pct

All Ranked Models

Max params:
Min confidence:
30 of 30
RankModelScore
#1gemini-2.5-pro

Strong on FACTS Benchmark Suite facts_grounding_score_pct (100%) and Vectara HHEM Leaderboard overall_hallucination_error_pct (76%)

27.0%
#2gpt-4.1-20250414

Strong on Vectara HHEM Leaderboard overall_hallucination_error_pct (82%) and Galileo Agent Leaderboard v2 Insurance AC (85%)

26.8%
#3gemini-3-pro-preview

Strong on FACTS Benchmark Suite facts_grounding_score_pct (88%) and FACTS Benchmark Suite facts_search_score_pct (100%)

24.1%
#4Grok-4-0709
20.7%
#5gpt-5-mini-2025-08-07
19.6%
#6gemini-2.5-flash
19.4%
#7gpt-5-2025-08-07
19.2%
#8claude-sonnet-4-20250514
18.3%
#9anthropic/claude-sonnet-4.6
18.1%
#10google/gemini-3.1-pro-preview
16.6%
#11openai/gpt-5.4-2026-03-05
16.3%
#12claude-opus-4-5-20251101
15.8%
#13gpt-4.1-mini-20250414
15.2%
#14gpt-5.1-2025-11-13
14.5%
#15gemini-3-flash-preview
14.1%
#16google/gemini-3.1-flash-lite-preview
13.5%
#17xai-org/grok-4-fast-reasoning
13.3%
#18xai-org/grok-4-1-fast-reasoning
12.7%
#19anthropic/claude-opus-4-6-thinking
12.6%
#20gpt-5.2-2025-12-11
12.6%
#21kimi/kimi-k2.5-thinking
12.0%
#22anthropic/claude-opus-4-5-20251101-thinking
11.7%
#24qwen-2.5-72b-instruct
11.1%
#25anthropic/claude-sonnet-4-5-20250929-thinking
11.0%
#27zai/glm-5-thinking
10.7%
#28x-ai/grok-3
10.5%
#30o3-20250416
9.9%
#31mistralai/mistral-large-2512
9.9%
#33anthropic/claude-haiku-4-5-20251001-thinking
9.7%
#34alibaba/qwen3.5-flash
9.6%

Compare Models

Model A leads by +0.2%

Shareable Link →

Model A

gemini-2.5-pro

external/google/gemini-2-5-pro

27.0%

Rank #1

Confidence 41.6%28 evidence pts

FACTS Benchmark Suite: facts_grounding_score_pct

Value 100.0% · Conf 100.0% · Weight 2.6%

facts_benchmark_suite.facts_grounding_score_pct (Mar 12, 2026)

Vectara HHEM Leaderboard: overall_hallucination_error_pct

Value 76.0% · Conf 100.0% · Weight 2.1%

vectara_hhem_leaderboard.overall_hallucination_error_pct (Mar 12, 2026)

Vals CorpFin v2: overall_accuracy_pct

Value 78.4% · Conf 100.0% · Weight 1.6%

vals_corp_fin_v2.overall_accuracy_pct (Mar 12, 2026)

Galileo Agent Leaderboard v2: Insurance AC

Value 63.6% · Conf 100.0% · Weight 1.6%

galileo_agent_v2.insurance_ac (Mar 12, 2026)

Model B

gpt-4.1-20250414

external/openai/gpt-4-1-20250414

26.8%

Rank #2

Confidence 41.2%26 evidence pts

Vectara HHEM Leaderboard: overall_hallucination_error_pct

Value 82.5% · Conf 100.0% · Weight 2.2%

vectara_hhem_leaderboard.overall_hallucination_error_pct (Mar 12, 2026)

Galileo Agent Leaderboard v2: Insurance AC

Value 85.5% · Conf 100.0% · Weight 2.1%

galileo_agent_v2.insurance_ac (Mar 12, 2026)

Vals CorpFin v2: overall_accuracy_pct

Value 84.9% · Conf 100.0% · Weight 1.8%

vals_corp_fin_v2.overall_accuracy_pct (Mar 12, 2026)

OpenVLM OCRBench Official: ocrbench_score_pct

Value 87.7% · Conf 100.0% · Weight 1.5%

openvlm_ocrbench_official.ocrbench_score_pct (Mar 12, 2026)

Ranking Diagnostics & Missing Models

Source Lift

Ranked

51

Sources

8

Quality

Insufficient

Vals CorpFin v2

vals_corp_fin_v2

42 rows

1.3% avg lift

Vals Legal Bench

vals_legal_bench

36 rows

0.3% avg lift

Vals MedQA

vals_medqa

34 rows

0.3% avg lift

Vals LiveCodeBench

vals_lcb

34 rows

0.3% avg lift

Missing Strong Models

No obvious gaps right now.

Taxonomy Details

Core Tasks

task.multi_doc_synthesistask.risk_assessment

Required Modes

mode.long_context

Domains

domain.insurance_claims

Related Use Cases