BasedAGIBasedAGI
Menu
Rankings live

business_productivity

Cross-lingual summary

Summarize a document in one language into another language.

#1 Recommendation

gemini-2.5-flash

Strong on LanguageBench overall:mean (100%) and Galileo Agent Leaderboard v2 Avg TSQ (100%)

external/google/gemini-2-5-flash

25.1%

Score

32.6%

Confidence

Limited benchmark evidence for this use case.

53 ranked models with average evidence of 14.4 points. Rankings may shift as more benchmark data is ingested.

Ranked Models

30

Evidence Quality

81%

Scoring

Benchmark-backed

Top Signal

LanguageBench: overall:mean

All Ranked Models

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

Strong on LanguageBench overall:mean (100%) and Galileo Agent Leaderboard v2 Avg TSQ (100%)

25.1%
#2gemini-2.5-pro

Strong on Galileo Agent Leaderboard v2 Avg TSQ (79%) and FACTS Benchmark Suite facts_grounding_score_pct (100%)

23.0%
#3gemini-3-pro-preview

Strong on Vals Finance Agent overall_accuracy_pct (87%) and FACTS Benchmark Suite facts_grounding_score_pct (88%)

22.3%
#4gpt-4.1-20250414
21.5%
#5anthropic/claude-sonnet-4.6
18.2%
#6gpt-5-mini-2025-08-07
17.8%
#7Grok-4-0709
17.6%
#8google/gemini-3.1-pro-preview
17.2%
#9gpt-5-2025-08-07
16.8%
#11openai/gpt-5.4-2026-03-05
16.7%
#12claude-sonnet-4-20250514
15.9%
#13gpt-5.1-2025-11-13
14.7%
#14google/gemini-2.0-flash-001
14.7%
#16claude-opus-4-5-20251101
14.1%
#18google/gemini-3.1-flash-lite-preview
13.8%
#20gpt-5.2-2025-12-11
13.2%
#22anthropic/claude-opus-4-6-thinking
12.7%
#23xai-org/grok-4-fast-reasoning
12.1%
#24gemini-3-flash-preview
12.0%
#25anthropic/claude-opus-4-5-20251101-thinking
11.9%
#26kimi/kimi-k2.5-thinking
11.3%
#27anthropic/claude-sonnet-4-5-20250929-thinking
11.2%
#28xai-org/grok-4-1-fast-reasoning
11.2%
#33Llama-3.1-70B-Instruct
10.0%
#34phi-4
10.0%
#35anthropic/claude-opus-4-1-20250805
9.9%
#38anthropic/claude-haiku-4-5-20251001-thinking
9.5%
#40mistralai/mistral-large-2512
9.4%
#41gpt-4o-2024-08-06
9.3%
#42alibaba/qwen3.5-flash
9.2%

Compare Models

Model A leads by +2.1%

Shareable Link →

Model A

gemini-2.5-flash

external/google/gemini-2-5-flash

25.1%

Rank #1

Confidence 32.6%19 evidence pts

LanguageBench: overall:mean

Value 100.0% · Conf 100.0% · Weight 2.8%

languagebench.overall_mean (Mar 12, 2026)

Galileo Agent Leaderboard v2: Avg TSQ

Value 100.0% · Conf 100.0% · Weight 2.8%

galileo_agent_v2.avg_tsq (Mar 12, 2026)

LanguageBench Translation Official (Split): translation_to:bleu

Value 92.0% · Conf 100.0% · Weight 2.2%

languagebench_translation_official.translation_to_bleu (Mar 12, 2026)

LanguageBench Grammar/Clarity Official (Split): grammar_clarity_score_pct

Value 100.0% · Conf 100.0% · Weight 2.1%

languagebench_grammar_clarity_official.grammar_clarity_score_pct (Mar 12, 2026)

Model B

gemini-2.5-pro

external/google/gemini-2-5-pro

23.0%

Rank #2

Confidence 43.5%27 evidence pts

Galileo Agent Leaderboard v2: Avg TSQ

Value 79.5% · Conf 100.0% · Weight 2.3%

galileo_agent_v2.avg_tsq (Mar 12, 2026)

FACTS Benchmark Suite: facts_grounding_score_pct

Value 100.0% · Conf 100.0% · Weight 1.9%

facts_benchmark_suite.facts_grounding_score_pct (Mar 12, 2026)

Vals Mortgage Tax: overall_accuracy_pct

Value 99.0% · Conf 100.0% · Weight 1.6%

vals_mortgage_tax.overall_accuracy_pct (Mar 12, 2026)

Vectara HHEM Leaderboard: overall_hallucination_error_pct

Value 76.0% · Conf 100.0% · Weight 1.6%

vectara_hhem_leaderboard.overall_hallucination_error_pct (Mar 12, 2026)

Ranking Diagnostics & Missing Models

Source Lift

Ranked

53

Sources

8

Quality

Insufficient

Vals Legal Bench

vals_legal_bench

38 rows

0.4% avg lift

Vals MedQA

vals_medqa

37 rows

0.4% avg lift

Vals Tax Eval v2

vals_tax_eval_v2

37 rows

0.4% avg lift

Vals Mortgage Tax

vals_mortgage_tax

34 rows

1.4% avg lift

Missing Strong Models

gpt-4o

external/openai/gpt-4o

Rank #22

15.2%

Thin evidence after weighting
Taxonomy Details

Core Tasks

task.cross_lingual_summarization

Required Modes

mode.multilingualmode.long_context

Domains

domain.general_business

Related Use Cases