Best AI Models for Multilingual Tasks

AI models ranked by multilingual performance using MMLU benchmark scores across languages. Find the best LLM for translation and non-English tasks.

Last updated: 20m ago

#1 Model

GPT-5.4

Score: 95.4

Average Score

82.9

Across all ranked models

Models Ranked

With benchmark data

Weights:MMLU (60%)Arena Elo (40%)

Top Best for Multilingual Models by Weighted Score

Top 15 models by weighted score

LMMarketCap.com

Benchmark Breakdown

Per-benchmark scores for top 10 models

MMLU

Arena Elo

LMMarketCap.com

#	Model	Provider	Score	MMLU	Arena Elo
1	GPT-5.4OpenAI	OpenAI	95.4	94	97.5
2	Claude Opus 4.6Anthropic	Anthropic	95.3	92.1	100
3	Gemini 3.1 Pro PreviewGoogle	Google	95.2	92.6	99
4	GPT-5.2OpenAI	OpenAI	94.8	93.5	96.8
5	GPT-5.1OpenAI	OpenAI	94.3	93.2	95.8
6	GPT-5.5OpenAI	OpenAI	93.8	92.4	95.8
7	GPT-5OpenAI	OpenAI	93.5	93	94.2
8	Claude Sonnet 4.6Anthropic	Anthropic	92.1	91.2	93.3
9	Claude Sonnet 4.5Anthropic	Anthropic	91.3	90.8	92
10	Gemini 3 Flash PreviewGoogle	Google	91.1	88	95.7
11	Gemini 2.5 ProGoogle	Google	90.7	90.8	90.7
12	Claude Opus 4.5Anthropic	Anthropic	90.2	91.4	88.3
13	o3OpenAI	OpenAI	89.7	92.3	85.8
14	DeepSeek V3.2DeepSeek	DeepSeek	88.1	88.5	87.5
15	R1 0528DeepSeek	DeepSeek	86.9	91.5	80
16	Gemini 2.5 FlashGoogle	Google	84.5	85.8	82.5
17	Claude Opus 4.7Anthropic	Anthropic	83.7	--	98.5
18	DeepSeek V3 0324DeepSeek	DeepSeek	83.2	89.2	74.2
19	Claude Opus 4.8Anthropic	Anthropic	82	--	96.5
20	Gemini 3.5 FlashGoogle	Google	81.7	--	96.2
21	Llama 4 MaverickMeta	Meta	81.1	88	70.8
22	DeepSeek V3DeepSeek	DeepSeek	81	88.5	69.7
23	GPT-4.1OpenAI	OpenAI	80.2	89.2	66.7
24	GPT-4oOpenAI	OpenAI	79	88.7	64.3
25	GPT-5.5 ProOpenAI	OpenAI	78.5	92.4	--
26	Mistral LargeMistral AI	Mistral AI	76.2	84.7	63.3
27	Llama 3.3 70B InstructMeta	Meta	74.6	86.3	57.2
28	Claude Haiku 4.5Anthropic	Anthropic	73.4	84.5	56.7
29	Phi 4Microsoft	Microsoft	72.1	84.8	--
30	Llama 3.1 70B InstructMeta	Meta	71.5	86	49.7
31	GPT-4o-miniOpenAI	OpenAI	70.7	82	53.7
32	Llama 4 ScoutMeta	Meta	67.7	79.6	--
33	Gemma 2 27BGoogle	Google	63.9	75.2	--
34	Command ACohere	Cohere	51.1	--	60.2

How scores are calculated

Each model's score is a weighted average of its available benchmark results. When a model is missing some benchmarks, the weights are re-normalized across the benchmarks that are available. All scores are on a 0-100 scale. Data sourced from official model cards, published papers, and third-party evaluation platforms.

Other Specialty Leaderboards

Best for Coding Best for Math Best for Reasoning Best for Writing Best for Instructions Best for Data Analysis Best for Roleplay

Frequently Asked Questions

Based on our benchmark analysis, GPT-5.4 by OpenAI is currently the #1 ranked model for multilingual, with a weighted score of 95.4/100.

Models are ranked using a weighted average of MMLU, Arena Elo benchmark scores. All scores are normalized to a 0-100 scale.

We currently rank 34 models that have relevant benchmark data for multilingual tasks.