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: 10m ago

#1 Model

GPT-5.4

Score: 95.4

Average Score

73.1

Across all ranked models

Models Ranked

114

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	GPT-5.2OpenAI	OpenAI	94.8	93.5	96.8
4	GPT-5.1OpenAI	OpenAI	94.3	93.2	95.8
5	GPT-5OpenAI	OpenAI	93.5	93	94.2
6	Grok 4xAI	xAI	92.4	91.5	93.7
7	Claude Sonnet 4.6Anthropic	Anthropic	92.1	91.2	93.3
8	Claude Sonnet 4.5Anthropic	Anthropic	91.3	90.8	92
9	Gemini 3 Flash PreviewGoogle	Google	91.1	88	95.7
10	Gemini 2.5 ProGoogle	Google	90.7	90.8	90.7
11	Claude Opus 4.5Anthropic	Anthropic	90.2	91.4	88.3
12	o3OpenAI	OpenAI	89.7	92.3	85.8
13	Claude Opus 4Anthropic	Anthropic	89.3	91	86.7
14	R1 0528DeepSeek	DeepSeek	86.9	91.5	80
15	Claude Sonnet 4Anthropic	Anthropic	86.2	89.5	81.2
16	R1DeepSeek	DeepSeek	85.7	90.8	78.2
17	o1OpenAI	OpenAI	85.1	91.8	75
18	Gemini 2.5 FlashGoogle	Google	84.5	85.8	82.5
19	Claude 3.7 SonnetAnthropic	Anthropic	84.4	90.2	75.7
20	Gemini 3.1 Pro PreviewGoogle	Google	84.1	--	99
21	Claude Opus 4.7Anthropic	Anthropic	83.7	--	98.5
22	o3 MiniOpenAI	OpenAI	83.5	86.9	78.5
23	DeepSeek V3 0324DeepSeek	DeepSeek	83.2	89.2	74.2
24	Grok 3xAI	xAI	82.6	88.5	73.7
25	GPT-5.2 ChatOpenAI	OpenAI	81.7	--	96.2
26	GPT-5.5OpenAI	OpenAI	81.5	--	95.8
27	Llama 4 MaverickMeta	Meta	81.1	88	70.8
28	DeepSeek V3DeepSeek	DeepSeek	81	88.5	69.7
29	GLM 5.1Zhipu AI	Zhipu AI	80.9	--	95.2
30	Grok 4.1 FastxAI	xAI	80.3	--	94.5
31	GPT-4.1OpenAI	OpenAI	80.2	89.2	66.7
32	MiMo-V2.5-ProXiaomi	Xiaomi	79.9	--	94
33	DeepSeek V4 ProDeepSeek	DeepSeek	79.8	--	93.8
34	Kimi K2.6Moonshot AI	Moonshot AI	79.6	--	93.7
35	GPT-4oOpenAI	OpenAI	79	88.7	64.3
36	Qwen3.6 Max PreviewAlibaba	Alibaba	78.9	--	92.8
37	GLM 5Zhipu AI	Zhipu AI	78.9	--	92.8
38	Grok 4.3xAI	xAI	78.6	--	92.5
39	Gemma 4 31BGoogle	Google	78.1	--	91.8
40	Claude Opus 4.1Anthropic	Anthropic	77.8	--	91.5
41	Qwen3.6 PlusAlibaba	Alibaba	77.6	--	91.3
42	MiMo-V2-ProXiaomi	Xiaomi	77.5	--	91.2
43	Qwen3.5 397B A17BAlibaba	Alibaba	77.3	--	91
44	GLM 4.7Zhipu AI	Zhipu AI	76.9	--	90.5
45	Gemini 3.1 Flash Lite PreviewGoogle	Google	76.2	--	89.7
46	Gemma 4 26B A4B Google	Google	76.2	--	89.7
47	Mistral LargeMistral AI	Mistral AI	76.2	84.7	63.3
48	Gemini 2.0 FlashGoogle	Google	76	76.4	75.3
49	GPT-4 TurboOpenAI	OpenAI	75.6	86.5	59.2
50	DeepSeek V4 FlashDeepSeek	DeepSeek	75.5	--	88.8
51	Phi 4Microsoft	Microsoft	74.6	84.8	59.3
52	Llama 3.3 70B InstructMeta	Meta	74.6	86.3	57.2
53	GPT-5 ChatOpenAI	OpenAI	74.5	--	87.7
54	GLM 4.6Zhipu AI	Zhipu AI	74.5	--	87.7
55	DeepSeek V3.2DeepSeek	DeepSeek	74.2	--	87.3
56	DeepSeek V3.2 ExpDeepSeek	DeepSeek	74.1	--	87.2
57	MiMo-V2.5Xiaomi	Xiaomi	74.1	--	87.2
58	Grok 4 FastxAI	xAI	73.8	--	86.8
59	Qwen3.5-122B-A10BAlibaba	Alibaba	73.4	--	86.3
60	Hy3 previewTencent	Tencent	73.4	--	86.3
61	DeepSeek V3.1DeepSeek	DeepSeek	73.4	--	86.3
62	Claude Haiku 4.5Anthropic	Anthropic	73.4	84.5	56.7
63	DeepSeek V3.1 TerminusDeepSeek	DeepSeek	73.1	--	86
64	Qwen3 VL 235B A22B InstructAlibaba	Alibaba	73	--	85.8
65	GLM 4.5Zhipu AI	Zhipu AI	72.4	--	85.2
66	MiniMax M2.7MiniMax	MiniMax	71.8	--	84.5
67	Qwen3.5-27BAlibaba	Alibaba	71.7	--	84.3
68	Llama 3.1 70B InstructMeta	Meta	71.5	86	49.7
69	Qwen3 Next 80B A3B InstructAlibaba	Alibaba	71.1	--	83.7
70	GPT-4o-miniOpenAI	OpenAI	70.7	82	53.7
71	Qwen3.5-FlashAlibaba	Alibaba	70.6	--	83
72	Qwen3.5-35B-A3BAlibaba	Alibaba	70.4	--	82.8
73	Qwen3 VL 235B A22B ThinkingAlibaba	Alibaba	70.3	--	82.7
74	MiniMax M2.5MiniMax	MiniMax	70.1	--	82.5
75	Step 3.5 FlashStepFun	StepFun	69.8	--	82.2
76	GPT-5 MiniOpenAI	OpenAI	69.4	--	81.7
77	Claude 3.7 Sonnet (thinking)Anthropic	Anthropic	69	--	81.2
78	GPT-4.1 MiniOpenAI	OpenAI	68.3	--	80.3
79	Trinity Large Thinkingarcee-ai	arcee-ai	68	--	80
80	o4 MiniOpenAI	OpenAI	68	--	80
81	GLM 4.6VZhipu AI	Zhipu AI	67.7	--	79.7
82	Llama 4 ScoutMeta	Meta	67.7	79.6	--
83	Trinity Large Previewarcee-ai	arcee-ai	67.3	--	79.2
84	Claude 3.5 HaikuAnthropic	Anthropic	67.1	80.9	46.3
85	GLM 4.5 AirZhipu AI	Zhipu AI	67	--	78.8
86	Qwen3 Next 80B A3B ThinkingAlibaba	Alibaba	66.4	--	78.2
87	GLM 4.7 FlashZhipu AI	Zhipu AI	66.3	--	78
88	MiniMax M1MiniMax	MiniMax	65.6	--	77.2
89	o3 Mini HighOpenAI	OpenAI	65.6	--	77.2
90	Grok 3 Mini BetaxAI	xAI	64.7	--	76.2
91	GLM 4.5VZhipu AI	Zhipu AI	64.2	--	75.5
92	Gemini 2.0 Flash LiteGoogle	Google	64.2	--	75.5
93	gpt-oss-120bOpenAI	OpenAI	64.2	--	75.5
94	Gemma 2 27BGoogle	Google	63.9	75.2	--
95	Qwen3 8BAlibaba	Alibaba	63.3	--	74.5
96	Mercury 2Inception	Inception	63.3	--	74.5
97	MiniMax M2MiniMax	MiniMax	63.2	--	74.3
98	Llama 3.3 Nemotron Super 49B V1.5NVIDIA	NVIDIA	62.8	--	73.8
99	GPT-5 NanoOpenAI	OpenAI	61.9	--	72.8
100	Nova 2 LiteAmazon	Amazon	61.9	--	72.8
101	GPT-4.1 NanoOpenAI	OpenAI	59.8	--	70.3
102	GPT-4o-mini (2024-07-18)OpenAI	OpenAI	59.1	--	69.5
103	gpt-oss-20bOpenAI	OpenAI	59.1	--	69.5
104	Mistral Large 2407Mistral AI	Mistral AI	58.5	--	68.8
105	Olmo 3 32B ThinkAllen AI	Allen AI	57.4	--	67.5
106	Llama 3 70B InstructMeta	Meta	53.1	--	62.5
107	GPT-4OpenAI	OpenAI	53	--	62.3
108	Command ACohere	Cohere	51.1	--	60.2
109	Claude 3 HaikuAnthropic	Anthropic	51	--	60
110	Command R (08-2024)Cohere	Cohere	49.4	--	58.2
111	Llama 3 8B InstructMeta	Meta	45.6	--	53.7
112	Llama 3.1 8B InstructMeta	Meta	44.1	--	51.8
113	Llama 3.2 3B InstructMeta	Meta	37.7	--	44.3
114	Llama 3.2 1B InstructMeta	Meta	29.8	--	35

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 114 models that have relevant benchmark data for multilingual tasks.