LiT (Lost in Translation) Robustness - Idioms and Cultural Metaphors: leaderboard

LiT, Lost in Translation (2026), measures multilingual proficiency by round-trip translation: English passages pass through a sequence of four languages and back, and an LLM judge scores the result against the source with MQM error weights. Share of robustness sentences with idioms and cultural metaphors whose round trips score at least 80, averaged over eight translation sequences, in percent; higher is better.

Metric: MQM>=80 rate (%): share of round-trip translations whose back-translation to English scores at least 80 on an MQM error scale (minor -1, major -5, critical -25) from a Grok 4.1 Fast judge comparing it with the English source; each passage is translated serially through a four-language sequence and back, averaged over eight sequences (high, medium and low-resource languages); robustness subset: idioms and cultural metaphors; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 17 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1Gemini 3 FlashNon-reasoning65.6#175
2Qwen 3.5 397B A17B62.5#264
3GLM-554.2#269
4Kimi K2 (Thinking)49
5Gemma 4 31B (IT)Thinking47.9#350
6GLM-4.743.8#380
7Qwen 3 235B A22B 2507 (Thinking)42.7#466
8Qwen 3.5 35B A3B40.6#492
9GPT-OSS-120BHigh37.5#704
10DeepSeek V3.2 ExpNon-reasoning34.4#439
11MiniMax-M2.533.3#600
12Gemma 3 27B (IT)31.2#1324
13Qwen 3 235B A22B 2507 Instruct26#616
14Kimi K222.5#443
15MiMo-V2-FlashNon-reasoning15.6#684

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=lit-lost-in-translation-robustness-idioms-and-cultural-metaphors · How It Works · Data refreshed daily, snapshot 2026-11-21.