LLM Leaderboard

Ranked by Math Index · source AA

AA composite — MATH-500, AIME, etc.

IntelligenceAA LMArenaLMArena CodingAA MathAA MMLU-ProAA GPQAAA LiveCodeAA MATH-500AA AIMEAA HLEAA SciCodeAA

#1 by Math Index

GPT-5.2 (xhigh)

99.0

OpenAI · in $1.75 / out $14.0 per 1M · 77 tok/s

#	Model	Creator	Math	In $/1M	Out $/1M	tok/s	LMArena
1	GPT-5.2 (xhigh)	OpenAI	99.0	$1.75	$14.0	77	1411
2	GPT-5 Codex (high)	OpenAI	98.7	$1.25	$10.0	171	—
3	Gemini 3 Flash Preview (Reasoning)	Google	97.0	$0.50	$3.00	224	1466
4	GPT-5.2 (medium)	OpenAI	96.7	$1.75	$14.0	0	1411
5	DeepSeek V3.2 Speciale	DeepSeek	96.7	$0.00	$0.00	0	1424
6	MiMo-V2-Flash (Reasoning)	Xiaomi	96.3	$0.10	$0.30	0	—
7	Gemini 3 Pro Preview (high)	Google	95.7	$2.00	$12.0	0	1480
8	GPT-5.1 Codex (high)	OpenAI	95.7	$1.25	$10.0	178	1422
9	GLM-4.7 (Reasoning)	Z AI	95.0	$0.60	$2.20	114	1436
10	Kimi K2 Thinking	Kimi	94.7	$0.60	$2.50	130	1414
11	KAT-Coder-Pro V1	KwaiKAT	94.7	$0.00	$0.00	0	—
12	GPT-5 (high)	OpenAI	94.3	$1.25	$10.0	110	1405
13	Nova 2.0 Lite (high)	Amazon	94.3	$0.30	$2.50	179	—
14	GPT-5.1 (high)	OpenAI	94.0	$1.25	$10.0	106	1441
15	gpt-oss-120b (high)	OpenAI	93.4	$0.15	$0.60	292	1366
16	Grok 4	SpaceXAI	92.7	$5.50	$27.5	0	1452
17	DeepSeek V3.2 (Reasoning)	DeepSeek	92.0	$0.28	$0.42	0	1424
18	GPT-5 (medium)	OpenAI	91.7	$1.25	$10.0	82	—
19	GPT-5.1 Codex mini (high)	OpenAI	91.7	$0.25	$2.00	213	1422
20	Claude Opus 4.5 (Reasoning)	Anthropic	91.3	$5.00	$25.0	56	—
21	Qwen3 235B A22B 2507 (Reasoning)	Alibaba	91.0	$0.70	$8.40	62	1366
22	NVIDIA Nemotron 3 Nano 30B A3B (Reasoning)	NVIDIA	91.0	$0.05	$0.20	159	—
23	o4-mini (high)	OpenAI	90.7	$1.10	$4.40	180	—
24	GPT-5 mini (high)	OpenAI	90.7	$0.25	$2.00	103	1374
25	K-EXAONE (Reasoning)	LG AI Research	90.3	$0.00	$0.00	0	—
26	DeepSeek V3.1 Terminus (Reasoning)	DeepSeek	89.7	$1.64	$2.75	0	1419
27	Grok 4 Fast (Reasoning)	SpaceXAI	89.7	$0.20	$0.50	0	1398
28	Nova 2.0 Omni (medium)	Amazon	89.7	$0.30	$2.50	0	—
29	DeepSeek V3.1 (Reasoning)	DeepSeek	89.7	$0.59	$1.69	0	1419
30	Grok 4.1 Fast (Reasoning)	SpaceXAI	89.3	$0.00	$0.00	0	1408
31	Ring-1T	InclusionAI	89.3	$0.00	$0.00	0	—
32	gpt-oss-20b (high)	OpenAI	89.3	$0.05	$0.20	218	1288
33	Nova 2.0 Pro Preview (medium)	Amazon	89.0	$1.25	$10.0	130	—
34	Nova 2.0 Lite (medium)	Amazon	88.7	$0.30	$2.50	183	—
35	o3	OpenAI	88.3	$2.00	$8.00	135	—
36	Qwen3 VL 235B A22B (Reasoning)	Alibaba	88.3	$0.70	$8.40	61	—
37	Claude 4.5 Sonnet (Reasoning)	Anthropic	88.0	$3.00	$15.0	49	—
38	Apriel-v1.6-15B-Thinker	ServiceNow	88.0	$0.00	$0.00	0	—
39	INTELLECT-3	Prime Intellect	88.0	$0.00	$0.00	0	1356
40	Gemini 2.5 Pro	Google	87.7	$1.25	$10.0	145	1457
41	DeepSeek V3.2 Exp (Reasoning)	DeepSeek	87.7	$0.28	$0.42	0	1424
42	Apriel-v1.5-15B-Thinker	ServiceNow	87.5	$0.00	$0.00	0	—
43	Gemini 3 Pro Preview (low)	Google	86.7	$2.00	$12.0	0	1480
44	GLM-4.6 (Reasoning)	Z AI	86.0	$0.55	$2.20	55	1440
45	GLM-4.6V (Reasoning)	Z AI	85.3	$0.30	$0.90	59	1440
46	GPT-5 mini (medium)	OpenAI	85.0	$0.25	$2.00	102	—
47	ERNIE 5.0 Thinking Preview	Baidu	85.0	$0.00	$0.00	0	—
48	Grok 3 mini Reasoning (high)	SpaceXAI	84.7	$0.30	$0.50	49	—
49	Seed-OSS-36B-Instruct	ByteDance Seed	84.7	$0.21	$0.57	39	—
50	Qwen3 VL 32B (Reasoning)	Alibaba	84.7	$0.70	$8.40	98	—

Quality, pricing, and speed benchmarks via Artificial Analysis. Human preference Elo via LMArena. Refreshed daily at 06:00 UTC · last sync Sun, 19 Jul 2026 06:00:01 GMT.

About the metrics

Intelligence Index

Artificial Analysis composite of reasoning, knowledge, coding, math.

LMArena Elo

LMArena

Human preference Elo from chatbot arena head-to-head votes.

Coding Index

AA composite — LiveCodeBench, SciCode, etc.

Math Index

AA composite — MATH-500, AIME, etc.

MMLU-Pro

Multi-domain knowledge & reasoning. Harder than MMLU.

GPQA Diamond

Graduate-level science Q&A. Strong reasoning signal.

LiveCodeBench

Competitive programming, contamination-resistant.

MATH-500

Subset of the MATH benchmark.

AIME

American Invitational Mathematics Examination — olympiad math.

Humanity's Last Exam

Frontier reasoning benchmark; most models still score low.

SciCode

Scientific coding tasks.