Best AI Models for Reasoning

Q: Which AI model has the best reasoning?

Reasoning rankings shift as new models launch. The live leaderboard above shows current standings based on GPQA Diamond, ARC-AGI, and other reasoning benchmarks.

Q: What is GPQA Diamond?

GPQA Diamond is a graduate-level question answering benchmark designed to test deep reasoning. Questions are written by domain experts and verified to be answerable only with genuine understanding.

Q: How does ARC-AGI measure reasoning?

ARC-AGI tests abstract reasoning through visual pattern puzzles that require novel inference, not memorization. It is considered one of the hardest AI benchmarks.

AI models ranked by reasoning benchmarks. Compare GPQA Diamond, ARC-AGI, BBH, and other reasoning tests across all providers.

225

Models

Providers

110

Open Source

$0.57

Median $/1M in

Top 3

Full Rankings

#	Model	Avg	ARC-AGI	ARC-AGI-2	BBH	GPQA diamond	HLE	SimpleBench	$/1M in	Context
1	GPT-5.5 Pro🇺🇸 OpenAI	87.8	-	-	-	94.2	-	-	$30.00	400K
2	GPT-5.5🇺🇸 OpenAI	85.0	95.0	-	-	93.6	-	-	$5.00	400K
3	Claude Opus 5.5🇺🇸 Anthropic	82.2	97.5	92.5	-	87.5	-	-	$4.00	1.0M
4	Claude Mythos Preview🇺🇸 Anthropic	81.8	-	-	-	94.5	56.8	-	N/A	1.0M
5	GPT-6 Astra🇺🇸 OpenAI	76.7	98.5	95.0	-	94.4	52.5	-	$10.00	1.1M
6	DeepSeek-V2 (MoE-236B, May 2024)🇨🇳 DeepSeekOpen	76.5	-	-	71.7	-	-	-	N/A	0K
7	Claude Sonnet 5.5🇺🇸 Anthropic	75.0	-	-	-	94.1	-	-	$2.00	1.0M
8	U PaLM 2-L Unknown	73.4	-	-	-	-	-	-	N/A	0K
9	Claude Fable 5.1🇺🇸 Anthropic	71.3	97.5	90.0	-	-	43.8	-	$10.00	1.0M
10	Claude Opus 5🇺🇸 Anthropic	69.4	97.5	90.4	-	91.8	-	76.7	$5.00	1.0M
11	Claude Fable 5🇺🇸 Anthropic	68.1	98.5	89.2	-	81.1	-	78.3	$10.00	1.0M
12	phi-3-small 7.4B🇺🇸 MicrosoftOpen	67.4	-	-	72.1	-	-	-	N/A	0K
13	GPT-5.6 Terra🇺🇸 OpenAI	66.8	96.5	83.9	-	91.1	-	38.7	$2.00	1.1M
14	GPT-5.6 Sol🇺🇸 OpenAI	66.5	97.5	92.5	-	91.3	-	66.0	$2.00	1.1M
15	GPT-5.4 Pro🇺🇸 OpenAI	64.5	94.5	83.3	-	92.8	41.5	68.9	$30.00	1.1M
16	U PaLM 2-M Unknown	63.0	-	-	-	-	-	-	N/A	0K
17	DeepSeek V4 Flash🇨🇳 DeepSeekOpen	61.6	-	-	-	-	-	35.6	$0.03	1.0M
18	o3 Pro🇺🇸 OpenAI	61.3	59.3	4.9	-	-	-	-	$20.00	200K
19	Qwen3.7 Max🇨🇳 Alibaba QwenOpen	61.3	-	-	-	87.9	-	64.5	$1.25	1.0M
20	Qwen-14B🇨🇳 Alibaba QwenOpen	60.7	-	-	40.0	-	-	-	N/A	0K
21	DeepSeek V4 Pro 0813🇨🇳 DeepSeek	60.0	90.5	61.3	-	88.9	-	-	$0.55	1.0M
22	Qwen3.5 397B A17B🇨🇳 Alibaba QwenOpen	59.9	-	-	-	81.8	-	-	$0.55	262K
23	Kimi K3🇨🇳 moonshotai	59.6	94.5	60.4	-	90.8	-	52.8	$0.67	1.0M
24	U internlm-20b Unknown	59.0	-	-	36.7	-	-	-	N/A	0K
25	U PaLM 2-S Unknown	58.9	-	-	-	-	-	-	N/A	0K
26	phi-3-medium 14B🇺🇸 MicrosoftOpen	58.6	-	-	75.2	3.5	-	-	N/A	0K
27	Falcon 2 11B TIIOpen	58.0	-	-	-	-	-	-	N/A	0K
28	R1 0528🇨🇳 DeepSeekOpen	57.9	21.2	1.1	-	68.4	-	29.0	$0.50	164K
29	Mixtral 8x7B Instruct🇫🇷 Mistral AIOpen	57.8	-	-	-	7.5	-	-	$0.54	33K
30	GPT-5.6 Luna🇺🇸 OpenAI	57.7	88.0	59.5	-	88.8	-	36.2	$0.20	1.1M
31	Gemini 3.7 Flash🇺🇸 Google DeepMind	57.6	95.5	84.6	-	93.1	-	-	$0.75	1.0M
32	GPT-5.2 Pro🇺🇸 OpenAI	57.5	90.5	54.2	-	-	-	48.9	$21.00	400K
33	Grok 4.6🇺🇸 xAI	57.5	87.5	67.1	-	92.0	-	71.1	$2.00	500K
34	Gemini 3.8 Flash🇺🇸 Google DeepMind	57.4	-	-	-	93.9	41.7	-	$0.75	1.0M
35	Qwen3.6 27B🇨🇳 Alibaba QwenOpen	56.9	-	-	-	81.1	-	-	$0.32	262K
36	Claude Sonnet 5🇺🇸 Anthropic	56.7	-	-	-	87.4	-	52.7	$2.00	1.0M
37	GLM 5.2🇨🇳 z-aiOpen	56.4	77.0	22.8	-	89.1	-	50.6	$1.00	1.0M
38	Grok 4.3🇺🇸 xAI	56.4	-	-	-	85.1	-	-	$1.25	1.0M
39	Kimi K2.7 Code🇨🇳 moonshotaiOpen	56.2	-	-	-	83.8	-	49.5	$0.61	262K
40	Gemini 3 Pro🇺🇸 Google DeepMind	56.1	75.0	31.1	-	90.2	34.4	71.7	N/A	0K
41	DeepSeek V3🇨🇳 DeepSeekOpen	56.1	-	-	83.3	42.0	-	2.7	$0.26	164K
42	Kimi K2 0711🇨🇳 moonshotaiOpen	56.1	-	-	-	-	-	11.6	$0.57	131K
43	U Muse Spark Unknown	55.6	-	-	-	86.4	37.6	-	N/A	0K
44	Claude Opus 4.8🇺🇸 Anthropic	55.4	92.5	72.1	-	88.0	-	57.8	$5.00	1.0M
45	DeepSeek V4 Flash 0731🇨🇳 DeepSeek	55.4	89.0	61.4	-	88.0	-	53.3	$0.02	1.0M
46	GLM 5.3🇨🇳 z-ai	55.1	-	-	-	87.9	-	-	$1.40	1.0M
47	GPT-4 (older v0314)🇺🇸 OpenAI	55.0	-	-	-	14.3	-	-	$30.00	8K
48	Gemini 3.5 Flash🇺🇸 Google DeepMind	54.8	92.5	72.1	-	90.4	-	72.0	$1.50	1.0M
49	GLM 5🇨🇳 z-aiOpen	54.7	44.7	4.9	-	83.8	-	43.8	$0.60	205K
50	GLM 5.1🇨🇳 z-aiOpen	54.3	-	-	-	86.5	-	46.1	$0.97	205K
51	Mixtral 8x7B🇫🇷 Mistral AIOpen	54.3	-	-	-	7.5	-	-	N/A	0K
52	phi-3-mini 3.8B🇺🇸 MicrosoftOpen	54.3	-	-	62.3	-	-	-	N/A	0K
53	Grok 4.5🇺🇸 xAI	54.2	87.2	52.6	-	91.3	-	64.0	$2.00	500K
54	DeepSeek V4 Pro🇨🇳 DeepSeekOpen	53.9	-	-	-	87.9	-	-	$0.21	1.0M
55	Grok 4🇺🇸 xAI	53.9	66.7	16.0	-	82.7	-	52.6	N/A	0K
56	Qwen3 235B A22B🇨🇳 Alibaba QwenOpen	53.7	-	-	-	60.9	-	17.2	$0.46	131K
57	Gemini 3.6 Flash🇺🇸 Google DeepMind	53.7	91.2	60.4	-	92.2	-	-	$0.75	1.0M
58	Gemini 2.0 Pro🇺🇸 Google DeepMind	53.7	-	-	-	54.2	-	-	N/A	0K
59	Grok-3 mini🇺🇸 xAI	53.6	16.5	0.4	-	68.3	-	-	N/A	0K
60	Qwen3.6 Plus🇨🇳 Alibaba QwenOpen	53.5	-	-	-	84.5	-	-	$0.33	1.0M
61	U Nemotron-4 15B Unknown	53.4	-	-	44.9	-	-	-	N/A	0K
62	Grok 4 Fast🇺🇸 xAI	53.3	48.5	5.3	-	-	-	-	N/A	0K
63	Qwen2.5 Coder 32B Instruct🇨🇳 Alibaba QwenOpen	53.1	-	-	-	-	-	-	$0.66	33K
64	Qwen3 235B A22B Thinking 2507🇨🇳 Alibaba QwenOpen	53.1	-	-	-	73.4	-	-	$0.23	131K
65	gpt-oss-20b🇺🇸 OpenAIOpen	53.1	-	-	-	47.7	-	-	$0.02	131K
66	GPT-5🇺🇸 OpenAI	53.1	65.7	9.9	-	81.6	21.6	48.0	$1.25	400K
67	Gemma 4 31B🇺🇸 Google DeepMindOpen	52.8	-	-	-	67.7	-	-	$0.09	262K
68	Qwen3.8 Max (0902)🇨🇳 Alibaba Qwen	52.8	-	-	-	90.2	-	-	$2.00	1.0M
69	Qwen2.5-Coder-3B🇨🇳 Alibaba QwenOpen	52.2	-	-	-	-	-	-	N/A	0K
70	Kimi K2 Thinking🇨🇳 moonshotaiOpen	52.2	-	-	-	79.0	-	-	$0.60	262K
71	DeepSeek R1 Distill Qwen 14B🇨🇳 DeepSeekOpen	52.0	-	-	-	26.3	-	-	N/A	0K
72	GPT-5.4🇺🇸 OpenAI	51.7	93.7	74.0	-	91.1	33.0	-	$2.50	1.1M
73	Qwen3 32B🇨🇳 Alibaba QwenOpen	51.5	-	-	-	54.3	-	-	$0.08	131K
74	GPT-5 Mini🇺🇸 OpenAI	51.5	54.3	4.4	-	66.7	15.4	-	$0.25	400K
75	o3🇺🇸 OpenAI	51.2	60.8	6.5	-	75.8	16.3	43.7	$2.00	200K
76	GPT-4o (2024-05-13)🇺🇸 OpenAI	51.1	-	-	-	31.9	-	-	$5.00	128K
77	Qwen3.6 Max Preview🇨🇳 Alibaba QwenOpen	51.0	-	-	-	83.2	-	55.6	$1.03	262K
78	Qwen3.5-35B-A3B🇨🇳 Alibaba QwenOpen	50.9	-	-	-	78.0	-	-	$0.15	262K
79	Gemini 2.5 Pro🇺🇸 Google DeepMind	50.8	41.0	4.9	-	80.4	17.7	54.9	$1.25	1.0M
80	DeepSeek-Coder-V2-Lite-Base🇨🇳 DeepSeekOpen	50.8	-	-	-	-	-	-	N/A	0K
81	DeepSeek V3.1🇨🇳 DeepSeekOpen	50.7	-	-	-	-	-	28.0	$0.25	164K
82	Claude Opus 4.7🇺🇸 Anthropic	50.5	93.5	75.8	-	86.9	33.0	54.0	$5.00	1.0M
83	Qwen2.5 Coder 14B Instruct🇨🇳 AlibabaOpen	50.0	-	-	-	-	-	-	N/A	0K
84	MiniMax M3🇨🇳 minimaxOpen	49.9	-	-	-	87.9	-	35.0	$0.30	1.0M
85	Gemini 3 Flash Preview🇺🇸 Google DeepMind	49.6	84.7	33.6	-	85.9	-	53.3	$0.50	1.0M
86	Kimi K2.5🇨🇳 moonshotaiOpen	49.4	65.3	11.8	-	83.5	20.6	36.2	$0.45	262K
87	gpt-oss-120b🇺🇸 OpenAIOpen	49.3	-	-	-	67.7	-	6.5	$0.04	131K
88	Qwen2.5 72B Instruct🇨🇳 Alibaba QwenOpen	49.2	-	-	73.1	32.2	-	-	$0.36	33K
89	Grok 4.20🇺🇸 xAI	48.8	89.5	65.1	-	85.8	-	-	$1.25	2.0M
90	MiniMax M2.5🇨🇳 minimaxOpen	48.6	63.7	4.9	-	-	-	-	$0.27	205K
91	o4 Mini🇺🇸 OpenAI	48.6	58.7	6.1	-	72.8	13.9	26.4	$1.10	200K
92	Qwen3 235B A22B Instruct 2507🇨🇳 Alibaba QwenOpen	48.3	11.0	1.3	-	-	-	-	$0.09	262K
93	Gemini 3.1 Pro Preview🇺🇸 Google DeepMind	48.2	98.0	77.1	-	92.6	43.7	75.5	$2.00	1.0M
94	Claude 3.7 Sonnet🇺🇸 Anthropic	48.1	28.6	0.9	-	73.0	3.4	35.7	N/A	0K
95	Claude Opus 4.6🇺🇸 Anthropic	48.1	94.0	69.2	-	87.4	31.1	61.1	$5.00	1.0M
96	Gemini 2.0 Flash🇺🇸 Google DeepMind	48.0	-	1.3	-	52.2	-	17.3	$0.10	1.0M
97	U Stable Beluga 2 Unknown	47.8	-	-	59.1	-	-	-	N/A	0K
98	DeepSeek V3.2🇨🇳 DeepSeekOpen	47.7	57.0	4.0	-	77.9	-	-	$0.28	164K
99	Gemini 2.0 Flash (Dec 2024)🇺🇸 Google DeepMind	47.6	-	1.3	-	52.2	-	17.3	N/A	0K
100	GPT-5.2🇺🇸 OpenAI	47.0	86.2	52.9	-	88.5	24.2	35.0	$1.75	400K
101	GPT-5.1🇺🇸 OpenAI	47.0	72.8	17.6	-	83.5	19.8	43.8	$1.25	400K
102	Qwen3 Max🇨🇳 Alibaba QwenOpen	46.4	-	-	-	63.5	-	-	$0.78	262K
103	Nemotron 3 Ultra🇺🇸 NVIDIAOpen	46.3	-	-	-	80.5	-	-	$0.50	1.0M
104	Claude Opus 4🇺🇸 Anthropic	46.1	35.7	8.6	-	68.3	6.2	50.6	$15.00	200K
105	Gemma 4 26B A4B 🇺🇸 Google DeepMindOpen	46.0	-	-	-	64.3	-	-	$0.07	262K
106	GLM 4.7🇨🇳 z-aiOpen	46.0	-	-	-	77.8	-	37.2	$0.60	205K
107	Mistral Large 2411🇫🇷 Mistral AIOpen	45.8	-	-	-	35.1	-	-	$2.00	131K
108	Gemini 1.5 Flash (May 2024)🇺🇸 Google DeepMind	45.7	-	-	-	20.5	-	-	N/A	0K
109	R1🇨🇳 DeepSeekOpen	45.6	15.8	1.3	-	62.3	-	17.1	$0.70	64K
110	Claude Sonnet 4🇺🇸 Anthropic	45.5	40.0	5.9	-	72.3	3.1	34.6	$3.00	200K
111	Grok 3🇺🇸 xAI	45.5	5.5	0.0	-	67.7	-	23.3	N/A	0K
112	Qwen3 8B🇨🇳 Alibaba QwenOpen	45.3	-	-	-	42.3	-	-	$0.12	131K
113	o1🇺🇸 OpenAI	45.3	30.7	-	-	69.0	3.3	28.1	$15.00	200K
114	Claude Sonnet 4.6🇺🇸 Anthropic	45.3	86.5	60.4	-	83.2	-	-	$3.00	1.0M
115	Gemini 3.1 Flash Lite🇺🇸 Google DeepMind	44.5	-	-	-	75.8	4.0	-	$0.25	1.0M
116	DeepSeek R1 Distill Qwen 32B🇨🇳 DeepSeekOpen	44.4	-	-	-	52.2	-	-	N/A	0K
117	Falcon-180B TIIOpen	44.4	-	-	16.1	-	-	-	N/A	0K
118	Qwen2.5 Coder 7B Instruct🇨🇳 AlibabaOpen	44.4	-	-	-	-	-	-	N/A	0K
119	Qwen2.5 Coder 7B Instruct🇨🇳 Alibaba QwenOpen	44.4	-	-	-	-	-	-	$0.03	33K
120	Qwen3.6 Flash🇨🇳 Alibaba QwenOpen	44.2	-	-	-	77.8	-	22.2	$0.19	1.0M
121	GLM 5.3 Flash🇨🇳 z-ai	44.1	-	-	-	86.9	-	-	$0.15	1.0M
122	GPT-4o-mini (2024-07-18)🇺🇸 OpenAI	43.2	-	0.1	-	17.0	-	-	$0.15	128K
123	GPT-5 Pro🇺🇸 OpenAI	42.2	70.2	18.3	-	-	28.2	53.9	$15.00	400K
124	Gemma 3 27B (free)🇺🇸 Google DeepMindOpen	42.2	-	-	-	31.8	-	-	Free	131K
125	o1-preview🇺🇸 OpenAI	41.9	18.0	-	-	33.8	-	30.0	N/A	0K
126	Claude Opus 4.5🇺🇸 Anthropic	41.6	80.0	37.6	-	81.4	21.4	54.4	$5.00	200K
127	Mistral 7B V0.1🇫🇷 Mistral AIOpen	41.6	-	-	41.5	-	-	-	N/A	0K
128	Qwen3.6 35B A3B🇨🇳 Alibaba QwenOpen	41.6	-	-	-	79.8	-	-	$0.15	262K
129	Qwen3.7 Flash🇨🇳 Alibaba Qwen	41.5	-	-	-	76.4	-	-	$0.03	1.0M
130	Gemini 1.5 Pro (Feb 2024)🇺🇸 Google DeepMind	41.3	-	0.8	78.7	27.8	-	12.5	N/A	0K
131	Qwen3.7 Plus🇨🇳 Alibaba QwenOpen	41.2	-	-	-	83.8	-	-	$0.32	1.0M
132	Qwen2.5-Max🇨🇳 Alibaba QwenOpen	41.1	-	-	-	41.5	-	-	N/A	0K
133	Mistral Large🇫🇷 Mistral AIOpen	40.8	-	-	-	18.4	-	7.0	$2.00	128K
134	U Baichuan 2-7B Unknown	40.3	-	-	22.1	-	-	-	N/A	0K
135	Qwen3 14B🇨🇳 Alibaba QwenOpen	40.3	-	-	-	51.7	-	-	$0.12	131K
136	GPT-4.1🇺🇸 OpenAI	39.5	5.5	0.4	-	55.9	0.6	12.4	$2.00	1.0M
137	Claude 3.5 Sonnet🇺🇸 Anthropic	39.4	-	-	-	38.7	0.0	13.0	N/A	0K
138	Gemini 2.5 Flash🇺🇸 Google DeepMind	39.3	32.3	2.5	-	-	7.7	29.4	$0.30	1.0M
139	Mistral Medium 3🇫🇷 Mistral AIOpen	39.3	-	-	-	46.0	0.0	-	$0.40	131K
140	Mistral Small 3.1 24B🇫🇷 Mistral AIOpen	39.2	-	-	-	30.0	-	-	$0.35	128K
141	GPT-5 Nano🇺🇸 OpenAI	39.2	20.7	2.6	-	59.3	-	-	$0.05	400K
142	Mistral Large 2407🇫🇷 Mistral AIOpen	39.1	-	-	-	32.0	-	7.0	$2.00	131K
143	Phi 4🇺🇸 MicrosoftOpen	39.1	-	-	-	41.4	-	-	$0.07	16K
144	GPT-4 Turbo🇺🇸 OpenAI	39.0	-	-	66.8	7.5	-	10.1	$10.00	128K
145	Qwen2.5 Coder 1.5B Instruct🇨🇳 AlibabaOpen	38.8	-	-	-	-	-	-	N/A	0K
146	Qwen3 30B A3B🇨🇳 Alibaba QwenOpen	38.7	-	-	-	48.9	-	-	$0.12	131K
147	Qwen3.5-9B🇨🇳 Alibaba QwenOpen	38.5	-	-	-	72.0	-	-	$0.10	262K
148	Qwen2-72B🇨🇳 Alibaba QwenOpen	38.4	-	-	-	21.0	-	-	N/A	0K
149	Gemini 1.5 Pro (May 2024)🇺🇸 Google DeepMind	38.4	-	0.8	85.6	27.8	0.0	12.5	N/A	0K
150	Kimi K2.6🇨🇳 moonshotaiOpen	38.3	-	-	-	87.7	-	-	$0.95	262K
151	Llama 3.1 405B🇺🇸 MetaOpen	37.9	-	-	77.2	34.5	-	7.6	N/A	0K
152	Claude Sonnet 4.5🇺🇸 Anthropic	37.8	63.7	13.6	-	76.4	9.4	45.2	$3.00	1.0M
153	Gemini 2.0 Flash Thinking (Jan 2025)🇺🇸 Google DeepMind	37.7	-	-	-	42.8	1.9	16.8	N/A	0K
154	GPT-5.4 Mini🇺🇸 OpenAI	37.5	63.7	18.9	-	82.5	-	-	$0.75	400K
155	Claude 3.5 Haiku🇺🇸 Anthropic	36.7	-	-	-	17.5	-	-	N/A	0K
156	GPT-4.1 Mini🇺🇸 OpenAI	36.6	3.5	0.1	-	54.5	-	-	$0.40	1.0M
157	Gemma 3 27B🇺🇸 Google DeepMindOpen	36.6	-	-	-	31.8	-	-	$0.08	131K
158	Llama 3.1 70B Instruct🇺🇸 MetaOpen	36.3	-	-	-	25.6	-	-	$0.40	131K
159	Claude Opus 4.1🇺🇸 Anthropic	36.3	-	-	-	69.7	7.1	52.0	$15.00	200K
160	Gemini 3.5 Flash Lite🇺🇸 Google DeepMind	36.3	53.5	10.3	-	77.8	-	-	$0.30	1.0M
161	Llama 3.3 70B Instruct🇺🇸 MetaOpen	36.1	-	-	-	29.9	-	3.9	$0.10	131K
162	Llama 3.2 90B🇺🇸 MetaOpen	36.1	-	-	-	21.4	-	-	N/A	0K
163	Gemma 2 9B🇺🇸 Google DeepMindOpen	36.0	-	-	-	3.3	-	-	$0.03	8K
164	GPT-4.5🇺🇸 OpenAI	35.9	10.3	0.8	-	58.3	0.7	21.4	N/A	0K
165	Gemma 2 27B🇺🇸 Google DeepMindOpen	35.9	-	-	-	15.3	-	-	$0.65	8K
166	Qwen2.5 32B Instruct🇨🇳 AlibabaOpen	35.8	-	-	-	28.1	-	-	N/A	0K
167	GPT-4o (2024-08-06)🇺🇸 OpenAI	35.6	-	-	-	32.3	-	1.4	$2.50	128K
168	Llama 2-13B🇺🇸 MetaOpen	35.2	-	-	44.3	1.8	-	-	N/A	0K
169	o1-mini🇺🇸 OpenAI	35.0	14.0	0.8	-	49.8	-	1.7	N/A	0K
170	LLaMA-13B🇺🇸 MetaOpen	34.9	-	-	17.2	-	-	-	N/A	0K
171	Qwen 3.5 Plus (hosted 397B-A17B)🇨🇳 Alibaba QwenOpen	34.6	-	-	-	79.8	-	-	N/A	0K
172	Claude 2🇺🇸 Anthropic	34.3	-	-	-	12.9	-	-	N/A	0K
173	Grok-2 (Dec 2024)🇺🇸 xAI	34.3	-	-	-	38.4	-	7.2	N/A	0K
174	Qwen3.5-Flash🇨🇳 Alibaba QwenOpen	34.0	-	-	-	76.4	-	-	$0.07	1.0M
175	U XGen-7B Unknown	33.9	-	-	-	-	-	-	N/A	0K
176	Claude Haiku 4.5🇺🇸 Anthropic	33.6	47.7	4.0	-	61.6	-	-	$1.00	200K
177	Mistral Nemo🇫🇷 Mistral AIOpen	33.4	-	-	-	6.5	-	-	$0.02	131K
178	GPT-3.5 Turbo (older v0613)🇺🇸 OpenAI	32.8	-	-	48.8	2.9	-	-	$1.00	4K
179	GPT-5.4 Nano🇺🇸 OpenAI	32.7	51.5	5.7	-	71.3	-	-	$0.20	400K
180	U vicuna-13b-v1.1 Unknown	32.5	-	-	24.1	-	-	-	N/A	0K
181	o3 Mini🇺🇸 OpenAI	32.5	34.5	3.0	-	69.4	-	7.4	$1.10	200K
182	Llama 3 70B Instruct🇺🇸 MetaOpen	32.4	-	-	-	20.8	-	-	$0.51	8K
183	GPT-4o (2024-11-20)🇺🇸 OpenAI	32.3	4.5	0.0	-	30.5	0.0	1.4	$2.50	128K
184	Llama 3-70B🇺🇸 MetaOpen	32.1	-	-	-	20.8	-	-	N/A	0K
185	open_llama_7b🇺🇸 MetaOpen	32.0	-	-	-	-	-	-	N/A	0K
186	U MPT-30B Unknown	31.7	-	-	17.3	-	-	-	N/A	0K
187	QwQ 32B🇨🇳 Alibaba QwenOpen	30.9	-	-	-	53.8	-	-	$0.15	131K
188	GPT-4.1 Nano🇺🇸 OpenAI	30.9	0.1	0.1	-	31.9	-	-	$0.10	1.0M
189	Llama 3 8B Instruct🇺🇸 MetaOpen	30.8	-	-	-	1.4	-	-	$0.14	8K
190	Phi 2🇺🇸 MicrosoftOpen	30.2	-	-	45.9	-	-	-	N/A	0K
191	U Dolly 2.0-12b Unknown	29.2	-	-	-	-	-	-	N/A	0K
192	GPT-4o-mini🇺🇸 OpenAI	29.2	-	0.0	-	17.0	-	0.0	$0.15	128K
193	U Yi 6B UnknownOpen	29.2	-	-	29.6	0.0	-	-	N/A	0K
194	Gemma 2B🇺🇸 Google DeepMindOpen	29.1	-	-	13.6	-	-	-	N/A	0K
195	Llama 3.3 70B Instruct (free)🇺🇸 MetaOpen	29.1	-	-	-	29.9	-	3.9	Free	131K
196	Qwen-1_8B🇨🇳 Alibaba QwenOpen	28.7	-	-	4.3	-	-	-	N/A	0K
197	Claude 3 Opus🇺🇸 Anthropic	28.0	-	-	-	29.6	-	8.2	N/A	0K
198	Llama 4 Maverick🇺🇸 MetaOpen	27.9	4.4	0.1	-	56.0	0.9	13.2	$0.19	1.0M
199	Claude 3 Sonnet🇺🇸 Anthropic	27.5	-	-	-	20.8	-	-	N/A	0K
200	Command R (08-2024)🇨🇦 Cohere	25.8	-	-	-	-	-	0.9	$0.15	128K
201	Claude 3 Haiku🇺🇸 Anthropic	25.7	-	-	-	15.1	-	-	$0.25	200K
202	DeepSeek Coder 33B🇨🇳 DeepSeekOpen	25.4	-	-	-	-	-	-	N/A	0K
203	Qwen2.5 7B Instruct🇨🇳 Alibaba QwenOpen	24.6	-	-	-	14.0	-	-	$0.10	33K
204	U StarCoder 2 15B UnknownOpen	24.3	-	-	-	-	-	-	N/A	0K
205	Llama 3.1 8B Instruct🇺🇸 MetaOpen	24.2	-	-	-	2.6	-	-	$0.02	131K
206	Mistral Small 3.2 24B🇫🇷 Mistral AIOpen	23.9	-	-	-	32.1	-	-	$0.09	256K
207	Mixtral 8x22B Instruct🇫🇷 Mistral AIOpen	23.8	-	-	-	12.1	-	-	$2.00	66K
208	U Baichuan1-7B Unknown	23.7	-	-	10.0	-	-	-	N/A	0K
209	Cerebras-GPT-13B🇺🇸 OpenAI	23.4	-	-	-	-	-	-	N/A	0K
210	CodeQwen1.5-7B🇨🇳 Alibaba QwenOpen	23.1	-	-	-	-	-	-	N/A	0K
211	Qwen2.5 Coder 0.5B Instruct🇨🇳 AlibabaOpen	20.8	-	-	-	-	-	-	N/A	0K
212	Claude 2.1🇺🇸 Anthropic	20.5	-	-	-	10.6	-	-	N/A	0K
213	Mistral Small 3🇫🇷 Mistral AIOpen	20.3	-	-	-	29.7	-	-	$0.05	33K
214	U INTELLECT-1 Unknown	20.2	-	-	13.1	-	-	-	N/A	0K
215	Llama 4 Scout🇺🇸 MetaOpen	19.5	0.5	0.1	-	35.8	-	-	$0.10	1.3M
216	Gemini 1.0 Pro🇺🇸 Google DeepMind	18.8	-	-	-	11.9	-	-	N/A	0K
217	U RedPajama-INCITE-7B-Base Unknown	18.0	-	-	-	-	-	-	N/A	0K
218	DeepSeek Coder 6.7B🇨🇳 DeepSeekOpen	16.7	-	-	-	-	-	-	N/A	0K
219	U Magistral Small 1.1 Unknown	16.5	5.0	0.0	-	31.2	-	-	N/A	0K
220	Phi-1.5🇺🇸 MicrosoftOpen	16.3	-	-	-	-	-	-	N/A	0K
221	U stablelm-tuned-alpha-7b Unknown	13.6	-	-	-	-	-	-	N/A	0K
222	Gemma 3 12B🇺🇸 Google DeepMindOpen	11.2	-	-	-	19.3	-	-	$0.05	131K
223	Llama 3.2 1B Instruct🇺🇸 MetaOpen	9.4	-	-	-	0.0	-	-	$0.03	60K
224	Gemma 3 4B🇺🇸 Google DeepMindOpen	5.8	-	-	-	0.0	-	-	$0.05	131K
225	DeepSeek Coder 1.3B🇨🇳 DeepSeekOpen	3.2	-	-	-	-	-	-	N/A	0K

90+ Gold 80-89 70-79 60-69 <60Scores in % unless noted. Avg = unweighted mean across tested benchmarks.

About this category

Models ranked by reasoning performance across GPQA Diamond, ARC-AGI, BBH, and other logic and reasoning benchmarks. These tests measure abstract thinking, pattern recognition, and multi-step inference.

Related categories

Best AI Models for Math

AI models ranked by math benchmarks. Compare MATH-500, GSM8K, and competition-level math scores across all providers.

Best AI Models for Knowledge

AI models ranked by knowledge benchmarks. Compare MMLU-Pro, GPQA Diamond, SimpleQA, and other knowledge tests.

Flagship AI Models

The best AI model from each provider, ranked by benchmark score. Compare the flagships from OpenAI, Anthropic, Google, Meta, and more.

Frequently asked questions

Which AI model has the best reasoning?

Reasoning rankings shift as new models launch. The live leaderboard above shows current standings based on GPQA Diamond, ARC-AGI, and other reasoning benchmarks.

What is GPQA Diamond?

GPQA Diamond is a graduate-level question answering benchmark designed to test deep reasoning. Questions are written by domain experts and verified to be answerable only with genuine understanding.

How does ARC-AGI measure reasoning?

ARC-AGI tests abstract reasoning through visual pattern puzzles that require novel inference, not memorization. It is considered one of the hardest AI benchmarks.

Back to all models