Geopolitics

Narrative tracking

Track narratives across multi-lingual sources and flag contradictions.

task.multi_doc_synthesistask.contradiction_detection

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

Provisional leader

gemini-3.1-pro-preview

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

29.0%

Best benchmark score

32.9%

Confidence

All ranked models — top 3

🥇

gemini-3.1-pro-preview

29.0%

🥈

claude-sonnet-4

25.9%

🥉

gpt-5-2025-08-07

25.1%

Ranked Models

Evidence Quality

82%

Evidence Points

Top Signal

Vals Finance Agent: overall_accuracy_pct

All Ranked Models

30 of 30 models

Rank	Model	Score	Confidence	Price / 1M	Evidence sources
🥇	gemini-3.1-pro-preview Strong on Vals Finance Agent overall_accuracy_pct and FACTS Benchmark Suite facts_search_score_pct	29.0%	33%	$4.50	Vals Finance AgentFACTS Benchmark Suite
🥈	claude-sonnet-4 Strong on LanguageBench overall:mean and Vectara HHEM Leaderboard overall_hallucination_error_pct	25.9%	40%	$6.00	LanguageBenchVectara HHEM Leaderboard
🥉	gpt-5-2025-08-07 Strong on FACTS Benchmark Suite facts_grounding_score_pct and Vals Finance Agent overall_accuracy_pct	25.1%	34%	—	FACTS Benchmark SuiteVals Finance Agent
#4	gemini-2.5-pro Strong on FACTS Benchmark Suite facts_grounding_score_pct and Vectara HHEM Leaderboard overall_hallucination_error_pct	24.8%	45%	$3.44	FACTS Benchmark SuiteVectara HHEM Leaderboard
#5	gemini-3-pro-preview Strong on Vals Finance Agent overall_accuracy_pct and Vals CorpFin v2 overall_accuracy_pct	23.8%	32%	$4.50	Vals Finance AgentVals CorpFin v2
#6	gemini-2.5-flash Strong on LanguageBench overall:mean and FACTS Benchmark Suite facts_grounding_score_pct	23.7%	35%	$0.17	LanguageBenchFACTS Benchmark Suite
#7	gpt-5-mini-2025-08-07 Strong on Vals Finance Agent overall_accuracy_pct and Vals CorpFin v2 overall_accuracy_pct	22.9%	37%	—	Vals Finance AgentVals CorpFin v2
#8	gpt-5.2-2025-12-11 Strong on FACTS Benchmark Suite facts_grounding_score_pct and Vals CorpFin v2 overall_accuracy_pct	22.6%	28%	—	FACTS Benchmark SuiteVals CorpFin v2
#9	gemini-3-flash-preview Strong on Vals CorpFin v2 overall_accuracy_pct and FACTS Benchmark Suite facts_grounding_score_pct	22.1%	30%	$1.13	Vals CorpFin v2FACTS Benchmark Suite
#10	Grok-4-0709 Strong on Vals CorpFin v2 overall_accuracy_pct and Vals Finance Agent overall_accuracy_pct	21.7%	32%	—	Vals CorpFin v2Vals Finance Agent
#11	gemini-3.1-flash-lite-preview Strong on FACTS Benchmark Suite facts_grounding_score_pct and Vectara HHEM Leaderboard overall_hallucination_error_pct	20.6%	32%	$0.56	FACTS Benchmark SuiteVectara HHEM Leaderboard
#12	claude-sonnet-4.6 Strong on Vals Finance Agent overall_accuracy_pct and Vals CorpFin v2 overall_accuracy_pct	20.2%	26%	$6.00	Vals Finance AgentVals CorpFin v2
#13	gpt-4.1-20250414 Strong on Vectara HHEM Leaderboard overall_hallucination_error_pct and Vals CorpFin v2 overall_accuracy_pct	19.6%	31%	—	Vectara HHEM LeaderboardVals CorpFin v2
#14	gpt-5.4-2026-03-05 Strong on Vectara HHEM Leaderboard overall_hallucination_error_pct and Vals CorpFin v2 overall_accuracy_pct	19.0%	24%	—	Vectara HHEM LeaderboardVals CorpFin v2
#15	claude-opus-4-5-20251101 Strong on FACTS Benchmark Suite facts_grounding_score_pct and Vals CorpFin v2 overall_accuracy_pct	18.5%	28%	—	FACTS Benchmark SuiteVals CorpFin v2
#16	gpt-5.1-2025-11-13 Strong on Vals Finance Agent overall_accuracy_pct and Vals CorpFin v2 overall_accuracy_pct	17.1%	27%	—	Vals Finance AgentVals CorpFin v2
#17	grok-4-fast-reasoning Strong on Vals CorpFin v2 overall_accuracy_pct and Vals Finance Agent overall_accuracy_pct	17.0%	32%	$0.28	Vals CorpFin v2Vals Finance Agent
#18	o3-20250416 Strong on Vals CorpFin v2 overall_accuracy_pct and SciArena Leaderboard rating_elo	16.8%	27%	$3.50	Vals CorpFin v2SciArena Leaderboard
#19	grok-4-1-fast-reasoning Strong on Vals CorpFin v2 overall_accuracy_pct and Vals Finance Agent overall_accuracy_pct	15.3%	23%	$0.28	Vals CorpFin v2Vals Finance Agent
#20	kimi-k2.5-thinking Strong on Vals CorpFin v2 overall_accuracy_pct and Vals Finance Agent overall_accuracy_pct	14.5%	21%	—	Vals CorpFin v2Vals Finance Agent
#21	Claude-3.5-Sonnet Strong on LanguageBench overall:mean and LLM-AggreFact Leaderboard average_score_pct	14.2%	22%	$6.00	LanguageBenchLLM-AggreFact Leaderboard
#22	claude-opus-4-5-20251101-thinking Strong on Vals Finance Agent overall_accuracy_pct and Vals CorpFin v2 overall_accuracy_pct	14.1%	17%	—	Vals Finance AgentVals CorpFin v2
#23	claude-opus-4-6-thinking Strong on Vals CorpFin v2 overall_accuracy_pct and Vals Finance Agent overall_accuracy_pct	14.1%	15%	—	Vals CorpFin v2Vals Finance Agent
#25	grok-3 Strong on Vectara HHEM Leaderboard overall_hallucination_error_pct and Vals CorpFin v2 overall_accuracy_pct	13.1%	19%	$6.00	Vectara HHEM LeaderboardVals CorpFin v2
#26	claude-sonnet-4-5-20250929-thinking Strong on Vals Finance Agent overall_accuracy_pct and Vals CorpFin v2 overall_accuracy_pct	12.7%	17%	—	Vals Finance AgentVals CorpFin v2
#27	gpt-4.1 Strong on LanguageBench overall:mean and LanguageBench Translation Official (Split) translation_to:bleu	12.6%	17%	$3.50	LanguageBenchLanguageBench Translation Official (Split)
#30	grok-4.20-0309-reasoning Strong on Vals CorpFin v2 overall_accuracy_pct and Vals Finance Agent overall_accuracy_pct	11.9%	15%	—	Vals CorpFin v2Vals Finance Agent
#31	glm-5-thinking Strong on Vals CorpFin v2 overall_accuracy_pct and Vals Finance Agent overall_accuracy_pct	11.9%	17%	—	Vals CorpFin v2Vals Finance Agent
#32	gemini-2.0-flash-001 Strong on LanguageBench overall:mean and LanguageBench Translation Official (Split) translation_to:bleu	11.7%	18%	—	LanguageBenchLanguageBench Translation Official (Split)
#34	grok-4-1-fast-non-reasoning Strong on Vals Finance Agent overall_accuracy_pct and Vectara HHEM Leaderboard overall_answer_rate_pct	11.5%	22%	$0.28	Vals Finance AgentVectara HHEM Leaderboard

Compare Models

Select two different models above to compare their evidence side by side.

▶Ranking diagnostics & missing models

Source lift

Ranked

Sources

Quality

Low

Vals CorpFin v2

44 rows · 1.4% avg lift

Vals MedQA

35 rows · 0.3% avg lift

Vals Tax Eval v2

35 rows · 0.3% avg lift

Vals Legal Bench

34 rows · 0.3% avg lift

Missing frontier models

No obvious gaps right now.

▶Taxonomy & task details

Core tasks

task.multi_doc_synthesistask.contradiction_detection

Required modes

mode.long_context

Domains

domain.geopolitics_intel

Related in Geopolitics

Political risk brief

Summarize key developments into risks, scenarios, and actions.