Benchmark Almanac

NarrativeBench

Gemini 4 Argon 90.7% leads since November 2026
Not in model cards

Models seek, include, exclude, and recover narratives from multilingual news articles. Citation-based F1 or recall scores narrative coverage after checking the response language.

The release includes 1,333 articles, 460 queries, and 4,800 evaluation cases. Five tasks use different metrics, and language or framing gaps are lower-is-better.

Released
November 2026
Kind
Reading in many languages
Level
Everyday
Scored
Checked automatically
Answers
Open-ended
Tasks
All public
Measures
Capability
Language
Many languages
In and out
Text → Text
Interaction
Single turn
Items from
Real work and use
Refresh
One fixed set
Run by
Anyone
In views
Many languages
Strongest model tested
Gemini 4 Carbon #3 on the capability index
Last model added
November 2026
Scores refreshed
November 20th, 2026

In the authors’ words

Large language models (LLMs) increasingly mediate how people seek and consume information, including accounts that differ across languages and regions, especially on contested topics. Supporting diverse information seeking requires models to help users navigate competing narratives across languages and information streams. We define six capabilities necessary for diverse multilingual information seeking: three foundational capabilities for narrative inclusion, narrative exclusion, and missing narrative identification; and three higher-level capabilities for presenting diverse narratives, preserving narrative coverage under confirmatory queries, and maintaining information parity across languages. To guide the development of those information seeking capabilities, we introduce NarrativeBench, a benchmark comprising 1,333 news articles in 23 publication languages across 23 contested topics and 460 Reddit-grounded queries. Evaluating 14 LLMs across 24 query languages, we find that current models lack the capabilities needed to reliably navigate diverse narratives, limiting users' control over the perspectives they receive and leaving substantial gaps in exposure to competing accounts even when diversity is explicitly requested. Information access also differs across query framings and languages, risking echo chambers, linguistic filter bubbles, and information disparities that weaken the common ground needed for democratic discourse. NarrativeBench provides a foundation for developing information systems that give users across languages and regions agency to navigate diverse narratives.

Abstract of The Narrative Gap: Can LLMs Help Us Navigate Diverse Narratives Across Languages? (arXiv:2611.12750, CC0)
17
Models scored on its own leaderboard

Gemini 4 Argon leads with 90.7%

Best score per model from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart.

  1. 1 Gemini 4 Argon high 90.7%
  2. 2 Gemini 4 Hydrogen xhigh 88%
  3. 3 Gemini 4 Carbon high 78.6%
  4. 4 Gemini 3.8 Flash low 52.5%
  5. 5 GPT-6 Luna 26%
  6. 6 Gemini 2.5 Pro Jun 2025 25.2%
  7. 7 Gemini 2.5 Flash Sep 2025 21.8%
  8. 8 Gemma 4 26B A4B Reasoning disabled 21.4%
  9. 9 GPT-5.4 Mini 21%
  10. 10 DeepSeek-V4-Flash Reasoning disabled 20.4%
  11. 11 Kimi K2.6 19.8%
  12. 12 Qwen3.5-9B 18.8%
  13. 13 gpt-oss-120b low, 3K 16.5%
  14. 14 Cohere Command A 14.3%
  15. 15 gpt-oss-20b medium, 16K 14.3%
  16. 16 Qwen3.5-9B-Base 10.7%
  17. 17 Qwen3.5-0.8B 2.60%

Source: Official NarrativeBench leaderboard

#ModelScoreSettingDateSource
1 Gemini 4 Argon 90.7% high Sep 2026 Official leaderboard
2 Gemini 4 Hydrogen 88% xhigh Nov 2026 Official leaderboard
3 Gemini 4 Carbon 78.6% high Oct 2026 Official leaderboard
4 Gemini 3.8 Flash 52.5% low Sep 2026 Official leaderboard
5 GPT-6 Luna 26% Sep 2026 Official leaderboard
6 Gemini 2.5 Pro Jun 2025 25.2% Jun 2025 Official leaderboard
7 Gemini 2.5 Flash Sep 2025 21.8% Sep 2025 Official leaderboard
8 Gemma 4 26B A4B 21.4% Reasoning disabled Apr 2026 Official leaderboard
9 GPT-5.4 Mini 21% Mar 2026 Official leaderboard
10 DeepSeek-V4-Flash 20.4% Reasoning disabled Apr 2026 Official leaderboard
11 Kimi K2.6 19.8% Apr 2026 Official leaderboard
12 Qwen3.5-9B 18.8% Feb 2026 Official leaderboard
13 gpt-oss-120b 16.5% low, 3K Aug 2025 Official leaderboard
14 Cohere Command A 14.3% Mar 2025 Official leaderboard
15 gpt-oss-20b 14.3% medium, 16K Aug 2025 Official leaderboard
16 Qwen3.5-9B-Base 10.7% Feb 2026 Official leaderboard
17 Qwen3.5-0.8B 2.60% Feb 2026 Official leaderboard

Rankings by part

We read 5 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as Narrative inclusion, Narrative exclusion and Include Missing Narratives. Gemini 4 Hydrogen leads 3 of them, and 2 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.

Narrative inclusion

17 models we track have a score here.

  1. 1 Gemini 4 Argon 90.7%
  2. 2 Gemini 4 Hydrogen 88%
  3. 3 Gemini 4 Carbon 78.6%
  4. 4 Gemini 3.8 Flash 52.5%
  5. 5 GPT-6 Luna 26%
  6. 6 Gemini 2.5 Pro Jun 2025 25.2%
  7. 7 Gemini 2.5 Flash Sep 2025 21.8%
  8. 8 Gemma 4 26B A4B 21.4%
  9. 9 GPT-5.4 Mini 21%
  10. 10 DeepSeek-V4-Flash 20.4%

More effort adds a median 5 points

Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.

Lowest effort to highest, light to dark
Gemini 4 Argon
low
90.7%
Gemini 4 Hydrogen
low xhigh
88%
Gemini 4 Carbon
low
78.6%
gpt-oss-20b
low, 3K
14.3%

Source: Official NarrativeBench leaderboard

ModelEffortHarnessScoreCost per taskOutput tokens
Gemini 4 Argon high 90.7%
Gemini 4 Hydrogen xhigh 88%
Gemini 4 Argon low 85.3%
Gemini 4 Carbon high 78.6%
Gemini 4 Carbon low 78.1%
Gemini 4 Hydrogen low 72.4%
Gemini 3.8 Flash low 52.5%
GPT-6 Luna 26%
Gemini 2.5 Pro Jun 2025 25.2%
Gemini 2.5 Flash Sep 2025 21.8%
Gemma 4 26B A4B Reasoning disabled 21.4%
GPT-5.4 Mini 21%
DeepSeek-V4-Flash Reasoning disabled 20.4%
Kimi K2.6 19.8%
Qwen3.5-9B 18.8%
gpt-oss-120b low, 3K 16.5%
Cohere Command A 14.3%
gpt-oss-20b medium, 16K 14.3%
gpt-oss-20b low, 3K 12.9%
Qwen3.5-9B-Base 10.7%
Qwen3.5-0.8B 2.60%

Number one over time

Lab claims count from the day the card came out, independent runs from the model’s release.

Google DeepMind
Own leaderboard Passed 85% in Nov 2026, by Gemini 4 Argon

Source: Official NarrativeBench leaderboard

Gemini 4 Argon on the official leaderboard passed 85% in September 2026

Models released before it came out already scored that high. All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
On the benchmark’s own leaderboard Best so far
0% 20% 40% 60% 80% 100% 2025 2026 2027 Near the ceiling (85%) Gemini 4 Argon Gemini 3.8 Flash Gemini 2.5 Pro Cohere Command A 0% 20% 40% 60% 80% 100% 2025 2026 2027 Near the ceiling (85%) Gemini 4 Argon Gemini 3.8 Flash

Source: Official NarrativeBench leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Multilingual

All 545 in Multilingual
  • MMMLU In use MMLU professionally translated into 14 languages. Multiple-choice accuracy. Claude Mythos Preview 92.7%
  • WMT24++ In use English-to-55-language translation of WMT24 test data. Scored with automatic metrics such as MetricX or xCOMET. K-EXAONE 90.5%
  • MMLU-ProX In use MMLU-Pro translated into up to 29 languages with expert review. Multiple-choice accuracy per language. Qwen3.7-Max 87%
  • CLIcK In use Models answer 1,995 multiple-choice questions on Korean language and culture. Questions come from official Korean exams and textbooks in eleven categories. Solar Open 2 90.7%
  • Global MMLU Lite 3.0 In use Models answer 400 human-translated MMLU questions per language in 23 languages. Version 3.0 adds Czech, Hungarian, Oriya, Slovak and Tajik and updates Italian. Accuracy. Claude Opus 5 95.7%
  • MMMB In use Models answer multiple-choice questions about images in six languages across 15 categories. The set has about 12,000 questions. LFM2.5-VL-3B 83%

Data audit

All 9 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 3 pass 2 notes
  • Pass Independent runs Scores from the official leaderboard (17 entries).
  • Pass Models scored 17 models we track.
  • Pass Newest score The newest is from November 2026, for Gemini 4 Hydrogen.
  • Note Frontier models 2 of the top 10 have a score, which is fine for a retired test. The strongest tested is Gemini 4 Carbon (#3).
  • Note Headroom Saturated. The best is 90.7%, past 85% since September 2026.
Leaderboard import 3 pass
  • Pass Importer We read the board on isle-dev.github.io: 21 rows for models we track at the last read.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check