How is this ranking computed?

It is the same family composite as the leaderboard. Each benchmark is weighted by how much of its spread survives its own measurement error.

Best AI for math

Ranked by the mathematics composite: MATH level 5, FrontierMath and OTIS Mock AIME. Full leaderboard.

Rank	Model	Lab	Composite	Evidence	Out $/M
1	Claude Fable 5	Anthropic	100th	15 bench · 2 src	$50.00
2	AI Co-Mathematician	Google	100th	2 bench · 1 src	—
3	GPT-5.4 Pro	OpenAI	99th	1 bench · 1 src	—
4	GPT-6.1 Sol	OpenAI	99th	13 bench · 2 src	$10.00
5	GPT-6 Astra	OpenAI	98th	15 bench · 2 src	$50.00
6	Claude Opus 5.5	Anthropic	98th	15 bench · 2 src	$20.00
7	GPT-6 Astra	OpenAI	98th	1 bench · 1 src	—
8	GPT-6 Sol	OpenAI	97th	13 bench · 2 src	$10.00
9	Claude Fable 5.1	Anthropic	97th	14 bench · 2 src	$50.00
10	GPT-5.6 Sol	OpenAI	96th	2 bench · 1 src	$30.00
11	GPT-5.5 Pro pre-release	OpenAI	96th	4 bench · 1 src	—
12	GPT-5.5 Pro	OpenAI	96th	3 bench · 1 src	$180.00
13	Claude Sonnet 5.5	Anthropic	95th	13 bench · 2 src	$10.00
14	GPT-5.6 Sol	OpenAI	95th	13 bench · 2 src	$20.00
15	GPT-5.2 Pro	OpenAI	94th	1 bench · 1 src	—
16	Gemini 2.5 Pro Exp	Google	94th	2 bench · 1 src	—
17	Gemini 4 Argon	Google	94th	9 bench · 1 src	$20.00
18	GPT-5.4	OpenAI	93rd	14 bench · 2 src	$15.00
19	Claude Opus 5	Anthropic	93rd	10 bench · 2 src	$25.00
20	GPT-5.5	OpenAI	92nd	17 bench · 2 src	$30.00
21	Muse Wonder	Meta	92nd	6 bench · 1 src	$20.00
22	GPT-5.6 Terra	OpenAI	92nd	9 bench · 2 src	$12.00
23	GPT-5.4 Pro	OpenAI	91st	6 bench · 1 src	$180.00
24	Qwen3.7 Plus	Alibaba	91st	3 bench · 1 src	$1.60
25	Claude Opus 4.8	Anthropic	90th	14 bench · 2 src	$25.00
26	GPT-6 Luna	OpenAI	90th	12 bench · 2 src	$0.50
27	gpt-oss-120b	OpenAI	90th	5 bench · 3 src	free
28	Gemini 3 Pro Preview	Google	89th	5 bench · 2 src	$12.00
29	Nemotron 3 Ultra	NVIDIA	89th	3 bench · 1 src	$0.10
30	Qwen3.8 Max	Alibaba	88th	8 bench · 2 src	$6.00
31	Claude Haiku 5.5	Anthropic	88th	10 bench · 2 src	$0.50
32	Claude Opus 4.8	Anthropic	87th	3 bench · 1 src	—
33	Muse Spark 1.3	Meta	87th	6 bench · 2 src	$4.25
34	Claude Opus 4.7	Anthropic	87th	14 bench · 2 src	$25.00
35	Gemma 4 26B-A4B IT	Google	86th	2 bench · 1 src	$0.29
36	GPT-5.6 Luna	OpenAI	86th	9 bench · 2 src	$1.20
37	Muse Spark	Meta	85th	4 bench · 1 src	—
38	Gemini 3.7 Flash	Google	85th	9 bench · 2 src	$7.50
39	Kimi K3	Kimi	85th	9 bench · 2 src	$1.00
40	Claude Opus 4.6	Anthropic	84th	12 bench · 2 src	$25.00
41	Kimi K2.6	Kimi	84th	13 bench · 2 src	$4.00
42	Claude Sonnet 4.6	Anthropic	83rd	8 bench · 2 src	$15.94
43	GPT-5.2 Pro	OpenAI	83rd	3 bench · 1 src	$168.00
44	DeepSeek R1	DeepSeek	83rd	4 bench · 2 src	$2.60
45	Muse Spark Open	Meta	82nd	5 bench · 1 src	$4.25
46	Qwen3.8 Max (0902)	Alibaba	82nd	7 bench · 1 src	—
47	Gemini 3.1 Pro Preview	Google	81st	13 bench · 2 src	$12.00
48	Grok 4.6	xAI	81st	10 bench · 2 src	$6.00
49	Qwen3.7-Max	Alibaba	81st	10 bench · 2 src	$7.50
50	Claude Opus 4.6	Anthropic	80th	4 bench · 1 src	—
51	DeepSeek-R1-Distill-Llama-70B	DeepSeek	80th	3 bench · 1 src	$0.13
52	GPT 5.2	OpenAI	79th	13 bench · 3 src	$14.00
53	DeepSeek V4.1 Flash	DeepSeek	79th	9 bench · 2 src	$1.20
54	Gemma 4 31B IT	Google	79th	3 bench · 1 src	free
55	GLM-5.3	Zhipu	78th	8 bench · 2 src	$4.40
56	DeepSeek V4 Pro 0813	DeepSeek	78th	8 bench · 2 src	$3.96
57	Gemini 3.8 Flash	Google	77th	9 bench · 2 src	$7.50
58	DeepSeek-R1-Distill-Qwen-14B	DeepSeek	77th	3 bench · 1 src	$0.431
59	Gemini 2.5 Flash	Google	77th	2 bench · 2 src	$2.50
60	Gemini 3.5 Flash	Google	76th	14 bench · 3 src	$9.00
61	GLM-5.1	Zhipu	76th	8 bench · 1 src	$4.40
62	MiniMax M3	MiniMax	75th	5 bench · 2 src	$1.20
63	Claude Sonnet 5	Anthropic	75th	8 bench · 2 src	$10.00
64	GPT-5 Pro	OpenAI	75th	5 bench · 1 src	$120.00
65	Qwen3.6 Max Preview	Alibaba	74th	6 bench · 1 src	free
66	Mistral Large 4	Mistral	74th	9 bench · 2 src	$4.4156
67	DeepSeek-R1	DeepSeek	73rd	4 bench · 1 src	—
68	GPT 5	OpenAI	73rd	13 bench · 3 src	$10.00
69	GPT-5.6 Sol	OpenAI	73rd	4 bench · 1 src	—
70	Grok 4.5	xAI	72nd	8 bench · 2 src	$6.00
71	Qwen3.5 Plus	Alibaba	72nd	7 bench · 1 src	$2.21
72	GPT 5.1	OpenAI	71st	8 bench · 2 src	$10.00
73	Kimi K2.5	Kimi	71st	6 bench · 2 src	$2.80
74	seed-oss-36b-instruct	—	70th	2 bench · 1 src	$0.5978
75	DeepSeek V4 Flash 0731	DeepSeek	70th	8 bench · 2 src	$0.28
76	Qwen3-30B-A3B	Alibaba	70th	2 bench · 1 src	—
77	Gemini 2.5 Pro Preview	Google	69th	4 bench · 1 src	—
78	Qwen3.6 Plus	Alibaba	69th	9 bench · 2 src	$3.00
79	Qwen3 Max	Alibaba	68th	6 bench · 1 src	$6.00
80	Qwen3.6 27B	Alibaba	68th	4 bench · 2 src	$2.4754
81	GLM-5.2	Zhipu	68th	10 bench · 2 src	$4.40
82	Gemini 3 Flash Preview	Google	67th	10 bench · 2 src	$3.00
83	Grok 4.7	xAI	67th	9 bench · 2 src	$6.00
84	GLM-5.3-Flash	Zhipu	66th	7 bench · 2 src	$0.50
85	Gemini 3.6 Flash	Google	66th	9 bench · 2 src	$3.75
86	Gemini 2.5 Deep Think	Google	66th	2 bench · 1 src	—
87	Qwen3.5-35B-A3B	Alibaba	65th	2 bench · 1 src	$0.459
88	GPT-5.2	OpenAI	65th	4 bench · 1 src	—
89	Qwen3.5 9B	Alibaba	64th	2 bench · 1 src	—
90	Kimi K2.7 Code	Kimi	64th	7 bench · 2 src	$4.00
91	Kimi K2 Instruct	Kimi	64th	6 bench · 3 src	$2.5146
92	GPT-4.5	OpenAI	63rd	4 bench · 2 src	$150.00
93	Qwen3.5 397B-A17B	Alibaba	63rd	3 bench · 1 src	$1.032
94	QwQ 32B	Alibaba	62nd	3 bench · 2 src	$1.00
95	Qwen3.7 Plus	Alibaba	62nd	4 bench · 1 src	—
96	Grok 4	xAI	62nd	5 bench · 2 src	$15.00
97	glm-4.7-flash	—	61st	2 bench · 1 src	$0.40
98	GPT 5 mini	OpenAI	61st	11 bench · 2 src	$2.00
99	Grok 4.20	xAI	60th	6 bench · 1 src	$12.00
100	Inkling Small	Thinking Machines	60th	6 bench · 1 src	$1.20
101	o3	OpenAI	60th	11 bench · 3 src	$8.00
102	GPT-5.5	OpenAI	59th	4 bench · 1 src	—
103	GPT-5.4	OpenAI	59th	4 bench · 1 src	—
104	Qwen3.5 397B-A17B	Alibaba	58th	4 bench · 1 src	—
105	Grok-3 mini	xAI	58th	5 bench · 1 src	$0.50
106	Grok 4.3 Beta	xAI	58th	6 bench · 1 src	—
107	Gemini 2.0 Pro	Google	57th	3 bench · 1 src	$4.998
108	Qwen3 8B	Alibaba	57th	2 bench · 1 src	free
109	Qwen3.5-4B	Alibaba	56th	1 bench · 1 src	$0.07
110	Qwen3-235B-A22B	Alibaba	56th	6 bench · 1 src	—
111	DeepSeek-R1-Distill-Qwen-32B	DeepSeek	56th	2 bench · 1 src	$0.861
112	DeepSeek-V3.2-Exp	DeepSeek	55th	6 bench · 3 src	$0.89
113	GPT-5.4 nano	OpenAI	55th	10 bench · 2 src	$1.25
114	Claude Haiku 4.5	Anthropic	54th	7 bench · 1 src	$5.00
115	GPT-5.4 mini	OpenAI	54th	10 bench · 2 src	$4.50
116	Claude Opus 4	Anthropic	53rd	7 bench · 2 src	—
117	o1-mini	OpenAI	53rd	5 bench · 2 src	$4.40
118	Claude Sonnet 4	Anthropic	53rd	7 bench · 3 src	—
119	DeepSeek v4	DeepSeek	52nd	7 bench · 1 src	—
120	Grok 3 Beta	xAI	52nd	6 bench · 2 src	—
121	GPT-5.6 Terra	OpenAI	51st	4 bench · 1 src	—
122	o4-mini	OpenAI	51st	12 bench · 3 src	$4.00
123	Gemini 2.0 Flash Thinking	Google	51st	6 bench · 3 src	$0.42
124	qwen3-4b-instruct-2507	—	50th	2 bench · 1 src	—
125	Qwen3.7 Flash	Alibaba	50th	4 bench · 1 src	$0.13
126	Claude 3.7 Sonnet	Anthropic	49th	7 bench · 3 src	$15.00
127	Claude Sonnet 4.5	Anthropic	49th	10 bench · 1 src	$15.00
128	GLM 5	Zhipu	49th	5 bench · 2 src	$3.20
129	Qwen3.6 27B	Alibaba	48th	4 bench · 1 src	—
130	gpt-oss-20b	OpenAI	48th	2 bench · 1 src	$0.30
131	Gemma 3 27B	Google	47th	4 bench · 2 src	free
132	Qwen3.6 35B-A3B	Alibaba	47th	3 bench · 1 src	free
133	Claude Opus 4.5	Anthropic	47th	6 bench · 1 src	—
134	GLM-4.7	Zhipu	46th	6 bench · 1 src	$2.20
135	Claude Opus 4.5	Anthropic	46th	12 bench · 2 src	$25.00
136	o1-2024-12-17	OpenAI	45th	9 bench · 3 src	$54.00
137	Qwen3	Alibaba	45th	3 bench · 1 src	—
138	Mistral Medium 3	Mistral	45th	4 bench · 1 src	$7.50
139	Qwen3.7 Flash	Alibaba	44th	4 bench · 1 src	—
140	DeepSeek-V3	DeepSeek	44th	4 bench · 1 src	—
141	Inkling	Thinking Machines	43rd	7 bench · 2 src	$4.05
142	DeepSeek v4	DeepSeek	43rd	4 bench · 1 src	—
143	Qwen3-14B	Alibaba	43rd	2 bench · 1 src	$0.24
144	GPT 5 nano	OpenAI	42nd	11 bench · 2 src	$0.40
145	O3 Mini	OpenAI	42nd	12 bench · 3 src	$4.40
146	Qwen3.6 Flash	Alibaba	41st	8 bench · 1 src	$1.50
147	Qwen3 32B	Alibaba	41st	3 bench · 2 src	—
148	Gemini 2.5 Pro	Google	41st	7 bench · 2 src	$10.00
149	Qwen3-30B-A3B	Alibaba	40th	2 bench · 1 src	$0.50
150	DeepSeek-R1-0528	—	40th	2 bench · 1 src	$0.20
151	Qwen3.6 35B-A3B	Alibaba	39th	4 bench · 1 src	—
152	GPT 4.1 mini	OpenAI	39th	10 bench · 3 src	$1.60
153	Qwen Plus	Alibaba	39th	3 bench · 1 src	—
154	Qwen3.5 9B	Alibaba	38th	2 bench · 1 src	$0.13
155	GPT-4.1 nano	OpenAI	38th	7 bench · 2 src	$0.40
156	GPT-5.6 Luna	OpenAI	37th	5 bench · 1 src	—
157	Phi-4	Microsoft	37th	3 bench · 1 src	$0.50
158	Grok 4 Heavy	xAI	36th	1 bench · 1 src	—
159	Llama 4 Maverick Instruct	Meta	36th	6 bench · 3 src	$0.60
160	Qwen3.5 Flash	Alibaba	36th	8 bench · 1 src	$0.368
161	GLM-5.2	Zhipu	35th	5 bench · 1 src	—
162	GPT 4.1	OpenAI	35th	10 bench · 3 src	$8.00
163	Qwen2-72B	Alibaba	34th	2 bench · 1 src	$0.90
164	Gemini 2.5 Flash	Google	34th	2 bench · 1 src	—
165	Qwen2.5-72B	Alibaba	34th	3 bench · 1 src	$0.40
166	Qwen2.5-Max	Alibaba	33rd	4 bench · 1 src	—
167	GPT-4 Turbo	OpenAI	33rd	2 bench · 1 src	—
168	Gemini 1.5 Pro	Google	32nd	3 bench · 1 src	—
169	DeepSeek-V3	DeepSeek	32nd	4 bench · 1 src	$1.10
170	GPT-5.5 Instant	OpenAI	32nd	4 bench · 1 src	—
171	Qwen3.6 Plus	Alibaba	31st	2 bench · 1 src	—
172	Claude Sonnet 4.5	Anthropic	31st	8 bench · 1 src	—
173	Grok-2	xAI	30th	4 bench · 1 src	$10.00
174	Gemini 3.1 Flash-Lite	Google	30th	3 bench · 1 src	$1.50
175	Qwen2.5-32B	Alibaba	30th	3 bench · 1 src	$0.90
176	Qwen Turbo	Alibaba	29th	3 bench · 1 src	$0.09
177	Llama 4 Scout Instruct	Meta	29th	5 bench · 2 src	$0.30
178	Claude 3.5 Sonnet	Anthropic	28th	5 bench · 1 src	—
179	Gemini 3.5 Flash-Lite	Google	28th	7 bench · 2 src	$2.50
180	GLM 4.6	Zhipu	28th	3 bench · 2 src	$2.20
181	GPT-5.1	OpenAI	27th	5 bench · 1 src	—
182	Claude Opus 4.1	Anthropic	27th	9 bench · 1 src	$75.00
183	Mistral Small 3.2	Mistral	26th	2 bench · 1 src	—
184	Llama 3.1-405B	Meta	26th	3 bench · 1 src	—
185	Claude 3.5 Sonnet	Anthropic	26th	5 bench · 1 src	—
186	GLM 4.5	Zhipu	25th	3 bench · 2 src	$2.20
187	Magistral Small 1.2	Mistral	25th	2 bench · 1 src	—
188	MiniMax M3	MiniMax	24th	4 bench · 1 src	—
189	Mistral Large 2	Mistral	24th	4 bench · 1 src	—
190	Magistral Small 1.0	Mistral	24th	3 bench · 1 src	—
191	Mistral Small 3.1	Mistral	23rd	3 bench · 1 src	—
192	Mistral Small 3	Mistral	23rd	3 bench · 1 src	$0.08
193	Gemini 1.5 Flash	Google	22nd	4 bench · 1 src	—
194	glm-4.7-flash_none	—	22nd	2 bench · 1 src	—
195	Claude 3.5 Haiku	Anthropic	22nd	4 bench · 1 src	—
196	Qwen Plus	Alibaba	21st	1 bench · 1 src	—
197	Tülu 3 70B	Ai2	21st	3 bench · 1 src	—
198	Llama 3.3 70B	Meta	20th	3 bench · 1 src	$0.72
199	WizardLM-2 8x22B	Microsoft	20th	2 bench · 1 src	$0.48
200	GPT-5.4 nano	OpenAI	19th	5 bench · 1 src	—
201	Qwen3 8B	Alibaba	19th	2 bench · 1 src	—
202	Yi-1.5-34B	01.AI	19th	2 bench · 1 src	—
203	Qwen3.6 Flash	Alibaba	18th	2 bench · 1 src	—
204	DeepSeek-R1-Distill-Qwen-1.5B	DeepSeek	18th	2 bench · 1 src	free
205	Mixtral 8x22B	Mistral	17th	2 bench · 1 src	$1.20
206	Llama 3.2 90B	Meta	17th	3 bench · 1 src	$0.72
207	GPT-5.4 mini	OpenAI	17th	5 bench · 1 src	—
208	Claude 3 Opus	Anthropic	16th	6 bench · 2 src	$75.00
209	Llama 3.1-70B	Meta	16th	3 bench · 1 src	$0.72
210	GPT 4o	OpenAI	15th	9 bench · 3 src	$10.00
211	GPT-4o mini	OpenAI	15th	8 bench · 2 src	$0.60
212	Gemma 3 12B	Google	15th	2 bench · 1 src	$0.50
213	Qwen3.5 Flash	Alibaba	14th	2 bench · 1 src	—
214	GPT-4 Turbo	OpenAI	14th	4 bench · 1 src	—
215	Gemma 2 27B	Google	13th	3 bench · 1 src	$0.65
216	phi-3-medium 14B	Microsoft	13th	2 bench · 1 src	—
217	Llama 3-70B	Meta	13th	3 bench · 1 src	$2.75
218	Mistral Large	Mistral	12th	3 bench · 1 src	$1.50
219	Hermes 2 Theta Llama-3 70B	Nous Research	12th	3 bench · 1 src	—
220	Llama 3.1-8B	Meta	11th	3 bench · 1 src	$0.22
221	GPT-4	OpenAI	11th	4 bench · 1 src	—
222	Ministral 8B	Mistral	11th	2 bench · 1 src	$0.15
223	Ministral 3B	Mistral	10th	2 bench · 1 src	$0.10
224	Gemma 2 9B	Google	10th	3 bench · 1 src	$0.20
225	Claude 3 Sonnet	Anthropic	9th	3 bench · 1 src	$15.00
226	Qwen3-1.7B	Alibaba	9th	2 bench · 1 src	$0.10
227	Gemma 3 4B	Google	9th	2 bench · 1 src	$0.10
228	DBRX	Databricks	8th	2 bench · 1 src	$1.20
229	Claude 3 Haiku	Anthropic	8th	3 bench · 1 src	$1.25
230	Mistral NeMo	Mistral	7th	2 bench · 1 src	$40.00
231	Gemini 1.5 Flash 8B	Google	7th	2 bench · 1 src	—
232	Claude 2	Anthropic	7th	4 bench · 2 src	—
233	Mixtral 8x7B	Mistral	6th	2 bench · 1 src	$0.50
234	Gemini 1.0 Pro	Google	6th	3 bench · 1 src	—
235	granite-4.0-micro	—	5th	2 bench · 1 src	—
236	Qwen2.5-7B	Alibaba	5th	2 bench · 1 src	$0.10
237	Claude 2.1	Anthropic	5th	2 bench · 1 src	—
238	Gemma 3 1B	Google	4th	2 bench · 1 src	—
239	GPT-3.5 Turbo	OpenAI	4th	5 bench · 1 src	$1.40
240	granite-4.0-1b	—	3rd	2 bench · 1 src	—
241	Llama 3-8B	Meta	3rd	3 bench · 1 src	$0.25
242	Llama 3.2 1B	Meta	2nd	2 bench · 1 src	$0.10
243	GPT-4	OpenAI	2nd	2 bench · 1 src	—
244	DeepSeek LLM 67B	DeepSeek	2nd	3 bench · 1 src	—
245	granite-4.0-350m	—	1st	2 bench · 1 src	—
246	Mistral 7B	Mistral	1st	3 bench · 1 src	$0.20
247	Yi-34B	01.AI	0th	2 bench · 1 src	$0.90
248	Llama 2-70B	Meta	0th	3 bench · 1 src	$2.75

Common questions

What is the best ai for math right now?: Ranked by the mathematics composite: MATH level 5, FrontierMath and OTIS Mock AIME.
How is this ranking computed?: It is the same family composite as the leaderboard. Each benchmark is weighted by how much of its spread survives its own measurement error.