LLM Leaderboard

Ranked by LMArena Elo · source LMArena

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

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

#1 by LMArena Elo

Gemini 4 Carbon (High)

1540

Google · in $4.00 / out $20.0 per 1M · 89 tok/s

#	Model	Creator	LMArena	In $/1M	Out $/1M	tok/s	LMArena
1	Gemini 4 Carbon (High)	Google	1540	$4.00	$20.0	89	1540
2	Gemini 4 Argon (High)	Google	1529	$4.00	$20.0	82	1529
3	Claude Opus 5.5 (High, Default Fallback)	Anthropic	1516	$4.00	$20.0	68	1516
4	Claude Fable 5.1 (Max, Default Fallback)	Anthropic	1514	$10.0	$50.0	68	1514
5	Claude Opus 5 (Max)	Anthropic	1508	$5.00	$25.0	0	1508
6	Claude Opus 5 (High)	Anthropic	1504	$5.00	$25.0	0	1504
7	Muse Wonder (Max)	Meta	1499	$4.00	$20.0	127	1499
8	Gemini 3.8 Flash (High)	Google	1498	$0.75	$3.75	129	1498
9	Claude Opus 4.6 (Non-reasoning, High)	Anthropic	1496	$5.00	$25.0	0	1496
10	Claude Opus 4.6 (Max)	Anthropic	1496	$5.00	$25.0	0	1496
11	Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 5 Fallback)	Anthropic	1493	$10.0	$50.0	0	1493
12	MiMo-V2.6-Pro	Xiaomi	1491	$0.43	$0.87	38	1491
13	Muse Spark 1.3 (Max)	Meta	1491	$1.25	$4.25	0	1491
14	Gemini 3.7 Flash (High)	Google	1489	$0.75	$3.75	0	1489
15	Qwen3.8 Max (0902)	Alibaba	1483	$2.00	$6.00	38	1483
16	Qwen3.8 Max	Alibaba	1483	$2.00	$6.00	0	1483
17	Claude Opus 4.7 (Non-reasoning, High)	Anthropic	1483	$5.00	$25.0	0	1483
18	Claude Opus 4.7 (Max)	Anthropic	1483	$5.00	$25.0	0	1483
19	Muse Spark 1.2 (Xhigh)	Meta	1483	$1.25	$4.25	0	1483
20	Gemini 3.5 Flash (High)	Google	1482	$1.50	$9.00	0	1482
21	Gemini 3.1 Pro Preview	Google	1480	$2.00	$12.0	130	1480
22	Gemini 3.6 Flash (High)	Google	1480	$0.75	$3.75	0	1480
23	Gemini 3 Pro Preview (Low)	Google	1479	$2.00	$12.0	0	1479
24	Gemini 3 Pro Preview (High)	Google	1479	$2.00	$12.0	0	1479
25	Muse Spark 1.1 (Xhigh)	Meta	1478	$1.25	$4.25	0	1478
26	Gemini 3.5 Flash (Medium)	Google	1478	$1.50	$9.00	0	1478
27	Kimi K3 (Max)	Kimi	1476	$3.00	$15.0	0	1476
28	Qwen3.7 Max	Alibaba	1475	$2.50	$7.50	0	1475
29	Muse Spark Open (Xhigh)	Meta	1474	$1.25	$4.25	183	1474
30	Muse Spark 1.3 (Xhigh)	Meta	1474	$1.25	$4.25	0	1474
31	Muse Spark	Meta	1474	$0.00	$0.00	0	1474
32	GLM-5.3 (Max)	Z AI	1474	$1.40	$4.40	73	1474
33	Muse Spark 1.3 (max)	Meta	1473	$1.25	$4.25	221	1473
34	GLM-5.3-Flash	Z AI	1473	$0.15	$0.50	63	1473
35	Claude Sonnet 5.5 (Xhigh, Default Fallback)	Anthropic	1473	$2.00	$10.0	92	1473
36	GPT-5.5 (High)	OpenAI	1472	$5.00	$30.0	0	1472
37	GLM-5.2 (Max)	Z AI	1470	$1.40	$4.40	0	1470
38	GPT-5.5 Instant (May 2026)	OpenAI	1467	$5.00	$30.0	0	1467
39	GPT-5.5 (Xhigh)	OpenAI	1467	$5.00	$30.0	0	1467
40	GPT-5.5 (Medium)	OpenAI	1467	$5.00	$30.0	0	1467
41	GPT-5.5 (Non-reasoning)	OpenAI	1467	$5.00	$30.0	0	1467
42	GPT-5.5 Pro (Xhigh)	OpenAI	1467	$0.00	$0.00	0	1467
43	GPT-5.5 Instant (June 2026)	OpenAI	1467	$5.00	$30.0	147	1467
44	GPT-5.5 (Low)	OpenAI	1467	$5.00	$30.0	0	1467
45	Gemini 3 Flash Preview (Non-reasoning)	Google	1466	$0.50	$3.00	0	1466
46	Gemini 3 Flash Preview (Reasoning)	Google	1466	$0.50	$3.00	0	1466
47	MiMo-V2.5-Pro (Non-reasoning)	Xiaomi	1464	$0.67	$1.85	0	1464
48	MiMo-V2.5-Pro (Reasoning)	Xiaomi	1464	$0.43	$0.87	0	1464
49	GLM-5.1 (Reasoning)	Z AI	1463	$1.19	$3.74	0	1463
50	GLM-5.1 (Non-reasoning)	Z AI	1463	$1.28	$4.07	0	1463

Quality, pricing, and speed benchmarks via Artificial Analysis. Human preference Elo via LMArena. Refreshed daily at 06:00 UTC · last sync Wed, 04 Nov 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.