BasedAGIBasedAGI
Security

Threat Intelligence Analysis

Analyzing threat reports, CVEs, and security advisories to produce structured risk assessments.

task.multi_doc_synthesistask.risk_assessment
Evidence quality is currently limited for this use case. Rankings below are useful for exploration, not a strong winner claim.

Provisional leader

gemini-2.5-pro

Current leader based on limited benchmark evidence. Treat this ranking as directional until coverage improves.

24.1%

Best benchmark score

37.7%

Confidence

Ranked Models

30

Evidence Quality

83%

Evidence Points

32

Top Signal

BaxBench Leaderboard: average_secure_pass_1_pct

Benchmark Sources

46

Last Updated

1d ago

Limited evidence. This use case needs coverage at 11% (need 12%) to reach sufficient quality. Scores below are directional — use with caution.

Benchmark Sources (8)

BaxBench Leaderboardaverage_secure_pass_1_pct · Apr 30, 2026
4% weight
VADER Leaderboardmean_score_pct · Mar 31, 2026
2% weight
Vectara HHEM Leaderboardoverall_hallucination_error_pct · Apr 30, 2026
2% weight
FACTS Benchmark Suitefacts_grounding_score_pct · Apr 30, 2026
2% weight
AI SOC LLM Leaderboardoverall_success_rate_pct · Apr 30, 2026
2% weight

All Ranked Models

30 of 30 models
RankModelScore
🥇gemini-2.5-pro

Strong on BaxBench Leaderboard average_secure_pass_1_pct and FACTS Benchmark Suite facts_grounding_score_pct

24.1%
🥈gpt-5-2025-08-07

Strong on BaxBench Leaderboard average_secure_pass_1_pct and FACTS Benchmark Suite facts_grounding_score_pct

19.5%
🥉o3-20250416

Strong on BaxBench Leaderboard average_secure_pass_1_pct and VADER Leaderboard mean_score_pct

18.8%
#4gpt-4.1-20250414

Strong on Vectara HHEM Leaderboard overall_hallucination_error_pct and Vals CorpFin v2 overall_accuracy_pct

18.1%
#5gemini-3.1-pro-preview

Strong on FACTS Benchmark Suite facts_search_score_pct and FACTS Benchmark Suite facts_grounding_score_pct

16.6%
#6claude-sonnet-4

Strong on BaxBench Leaderboard average_secure_pass_1_pct and Vectara HHEM Leaderboard overall_hallucination_error_pct

16.1%
#7gpt-5-mini-2025-08-07

Strong on Vals Finance Agent overall_accuracy_pct and Vals CorpFin v2 overall_accuracy_pct

15.7%
#8gpt-5.2-2025-12-11

Strong on FACTS Benchmark Suite facts_grounding_score_pct and Vals CorpFin v2 overall_accuracy_pct

15.2%
#9Grok-4-0709

Strong on Vals CorpFin v2 overall_accuracy_pct and Vals Finance Agent overall_accuracy_pct

15.0%
#10gemini-3-pro-preview

Strong on Vals CorpFin v2 overall_accuracy_pct and Berkeley Function Calling Leaderboard (Overall) Overall Acc

15.0%
#11gemini-2.5-flash

Strong on FACTS Benchmark Suite facts_grounding_score_pct and Vectara HHEM Leaderboard overall_hallucination_error_pct

14.3%
#12gemini-3-flash-preview

Strong on Vals CorpFin v2 overall_accuracy_pct and FACTS Benchmark Suite facts_grounding_score_pct

14.3%
#13qwen-2.5-72b-instruct

Strong on Open LLM Leaderboard MMLU-Pro mmlu_pro_accuracy_pct and Open LLM Leaderboard GPQA gpqa

13.7%
#14claude-sonnet-4.6

Strong on Vals Finance Agent overall_accuracy_pct and Vals CorpFin v2 overall_accuracy_pct

12.8%
#15gemini-3.1-flash-lite-preview

Strong on FACTS Benchmark Suite facts_grounding_score_pct and Vectara HHEM Leaderboard overall_hallucination_error_pct

12.7%
#16claude-opus-4-5-20251101

Strong on FACTS Benchmark Suite facts_grounding_score_pct and Vals CorpFin v2 overall_accuracy_pct

12.6%
#17grok-4-1-fast-reasoning

Strong on Vals CorpFin v2 overall_accuracy_pct and Berkeley Function Calling Leaderboard (Overall) Overall Acc

12.3%
#18gpt-5.4-2026-03-05

Strong on Vectara HHEM Leaderboard overall_hallucination_error_pct and Vals CorpFin v2 overall_accuracy_pct

11.3%
#19phi-4

Strong on Open LLM Leaderboard GPQA gpqa and Open LLM Leaderboard MMLU-Pro mmlu_pro_accuracy_pct

11.2%
#20claude-opus-4.7

Strong on Vals Finance Agent overall_accuracy_pct and Vals CorpFin v2 overall_accuracy_pct

11.1%
#22grok-4-fast-reasoning

Strong on Vals CorpFin v2 overall_accuracy_pct and Vals Finance Agent overall_accuracy_pct

10.4%
#25gpt-5.1-2025-11-13

Strong on Vals CorpFin v2 overall_accuracy_pct and Vals Finance Agent overall_accuracy_pct

10.1%
#27grok-4-1-fast-non-reasoning

Strong on Berkeley Function Calling Leaderboard (Overall) Overall Acc and Vals Finance Agent overall_accuracy_pct

9.9%
#28Meta-Llama-3-8B-Instruct

Strong on LLM Trustworthy Leaderboard adv and LLM Trustworthy Leaderboard privacy

9.9%
#29Llama-3.3-70B-Instruct

Strong on Open LLM Leaderboard MMLU-Pro mmlu_pro_accuracy_pct and Open LLM Leaderboard IFEval ifeval

9.8%
#31Llama-3.1-70B-Instruct

Strong on Open LLM Leaderboard MMLU-Pro mmlu_pro_accuracy_pct and Open LLM Leaderboard GPQA gpqa

9.6%
#33gpt-4.1-mini-20250414

Strong on Vals CorpFin v2 overall_accuracy_pct and OpenVLM OCRBench Official ocrbench_score_pct

9.3%
#34MaziyarPanahi/calme-3.2-instruct-78b

Strong on Open LLM Leaderboard MMLU-Pro mmlu_pro_accuracy_pct and Open LLM Leaderboard GPQA gpqa

9.0%
#35Qwen2.5-32B-Instruct

Strong on Open LLM Leaderboard MMLU-Pro mmlu_pro_accuracy_pct and Open LLM Leaderboard GPQA gpqa

9.0%
#36MaziyarPanahi/calme-3.1-instruct-78b

Strong on Open LLM Leaderboard MMLU-Pro mmlu_pro_accuracy_pct and Open LLM Leaderboard GPQA gpqa

8.9%

Compare Models

Select two different models above to compare their evidence side by side.
Ranking diagnostics & missing models

Source lift

Ranked

4522

Sources

8

Quality

Low

Open LLM Leaderboard MMLU-Pro

4490 rows · 1.1% avg lift

Open LLM Leaderboard GPQA

4490 rows · 0.7% avg lift

Open LLM Leaderboard IFEval

4490 rows · 0.6% avg lift

Open LLM Leaderboard BBH

4490 rows · 0.5% avg lift

Missing frontier models

No obvious gaps right now.

Taxonomy & task details

Core tasks

task.multi_doc_synthesistask.risk_assessment

Required modes

mode.long_context

Domains

domain.cybersecurity_defense

Related in Security