ARC-AGI-1 Semi-Private Evaluation (ARC-AGI-1)

We show this table for reference; we do not rank on it.

Data verified Captured 2026-11-10

ARC Prize fluid-intelligence benchmark using novel visual grid transformations.

Semi-Private score on ARC-AGI-1 — Captured 2026-11-10

We mirror the published semi-private score view for ARC-AGI-1. Claude Fable 5 (Max) leads the public snapshot at 98.50%, followed by Claude Fable 5 (XHigh) (98.50%) and Claude Fable 5.2 (XHigh) (98.50%). We do not use these results to rank models overall.

1Closed

Claude Fable 5 (Max)

Anthropic

Semi-Private evaluation · max reasoning

98.50%

Overall 80.38Context 1M+

2Closed

Claude Fable 5 (XHigh)

Anthropic

Semi-Private evaluation · xhigh reasoning

98.50%

Overall 80.38Context 1M+

Claude Fable 5.2 (XHigh)

Anthropic

Semi-Private evaluation · xhigh reasoning

98.50%

298 configurationsReasoningCurrentDisplay onlyUpdated Captured 2026-11-10

Semi-Private score table (298 configurations)

Semi-Private score results for ARC-AGI-1
Rank	Model / configuration	Score	Parameters (B)	Open / closed
1	Claude Fable 5 (Max)AnthropicSemi-Private evaluation · max reasoning	98.50%	Not reported	Closed
2	Claude Fable 5 (XHigh)AnthropicSemi-Private evaluation · xhigh reasoning	98.50%	Not reported	Closed
3	Claude Fable 5.2 (XHigh)AnthropicSemi-Private evaluation · xhigh reasoning	98.50%	Not reported	Unknown
4	Claude Opus 5.5 (High)AnthropicSemi-Private evaluation · high reasoning	98.50%	Not reported	Closed
5	Gemini 3.8 Flash (High)GoogleSemi-Private evaluation · high reasoning	98.50%	Not reported	Closed
6	Gemini 4 Argon (High)GoogleSemi-Private evaluation · high reasoning	98.50%	Not reported	Closed
7	Gemini 4 Argon (Medium)GoogleSemi-Private evaluation · medium reasoning	98.50%	Not reported	Closed
8	Gemini 4 Carbon (Medium)GoogleSemi-Private evaluation · medium reasoning	98.50%	Not reported	Closed
9	GPT-6 Astra (High)OpenAISemi-Private evaluation · high reasoning	98.50%	Not reported	Closed
10	GPT-6 Astra (XHigh)OpenAISemi-Private evaluation · xhigh reasoning	98.50%	Not reported	Closed
11	GPT-6.1 Sol (High)OpenAISemi-Private evaluation · high reasoning	98.50%	Not reported	Closed
12	GPT-6.1 Sol (XHigh)OpenAISemi-Private evaluation · xhigh reasoning	98.50%	Not reported	Closed
13	Gemini 3.1 Pro (Preview)GoogleSemi-Private evaluation	98.00%	Not reported	Unknown
14	Claude Fable 5.1 (Max)AnthropicSemi-Private evaluation · max reasoning	97.50%	Not reported	Unknown
15	Claude Fable 5.2 (High)AnthropicSemi-Private evaluation · high reasoning	97.50%	Not reported	Unknown
16	Claude Fable 5.2 (Max)AnthropicSemi-Private evaluation · max reasoning	97.50%	Not reported	Unknown
17	Claude Fable 5.2 (Medium)AnthropicSemi-Private evaluation · medium reasoning	97.50%	Not reported	Unknown
18	Claude Opus 5 (High)AnthropicSemi-Private evaluation · high reasoning	97.50%	Not reported	Closed
19	Claude Opus 5 (Max)AnthropicSemi-Private evaluation · max reasoning	97.50%	Not reported	Closed
20	Claude Opus 5.5 (Max)AnthropicSemi-Private evaluation · max reasoning	97.50%	Not reported	Closed
21	Claude Opus 5.5 (Medium)AnthropicSemi-Private evaluation · medium reasoning	97.50%	Not reported	Closed
22	Claude Opus 5.5 (XHigh)AnthropicSemi-Private evaluation · xhigh reasoning	97.50%	Not reported	Closed
23	Gemini 3.8 Flash (Medium)GoogleSemi-Private evaluation · medium reasoning	97.50%	Not reported	Closed
24	Gemini 4 Carbon (High)GoogleSemi-Private evaluation · high reasoning	97.50%	Not reported	Closed
25	GPT-5.6 Sol (XHigh)OpenAISemi-Private evaluation · xhigh reasoning	97.50%	Not reported	Closed
26	GPT-6 Astra (Max)OpenAISemi-Private evaluation · max reasoning	97.50%	Not reported	Closed
27	GPT-6 Astra (Medium)OpenAISemi-Private evaluation · medium reasoning	97.50%	Not reported	Closed
28	GPT-5.6 Sol (High)OpenAISemi-Private evaluation · high reasoning	97.00%	Not reported	Closed
29	Claude Fable 5.1 (XHigh)AnthropicSemi-Private evaluation · xhigh reasoning	96.50%	Not reported	Unknown
30	Claude Sonnet 5.5 (Max)AnthropicSemi-Private evaluation · max reasoning	96.50%	Not reported	Unknown
31	Claude Sonnet 5.5 (XHigh)AnthropicSemi-Private evaluation · xhigh reasoning	96.50%	Not reported	Unknown
32	GPT-5.5 Pro (High)OpenAISemi-Private evaluation · high reasoning	96.50%	Not reported	Unknown
33	GPT-5.6 Sol (Max)OpenAISemi-Private evaluation · max reasoning	96.50%	Not reported	Closed
34	GPT-5.6 Terra (Max)OpenAISemi-Private evaluation · max reasoning	96.50%	Not reported	Closed
35	GPT-6 Astra (Low)OpenAISemi-Private evaluation · low reasoning	96.50%	Not reported	Closed
36	GPT-6.1 Sol (Max)OpenAISemi-Private evaluation · max reasoning	96.50%	Not reported	Closed
37	Inkling 2 Small (XHigh)Thinking MachinesSemi-Private evaluation · xhigh reasoning	96.50%	Not reported	Unknown
38	Claude Fable 5.1 (High)AnthropicSemi-Private evaluation · high reasoning	96.00%	Not reported	Unknown
39	Gemini 3 Deep Think (2/26)GoogleSemi-Private evaluation	96.00%	Not reported	Unknown
40	Claude Fable 5 (High)AnthropicSemi-Private evaluation · high reasoning	95.50%	Not reported	Closed
41	Gemini 3.7 Flash (High)GoogleSemi-Private evaluation · high reasoning	95.50%	Not reported	Closed
42	GPT-6 Sol (Max)OpenAISemi-Private evaluation · max reasoning	95.50%	Not reported	Closed
43	GPT-6.1 Sol (Medium)OpenAISemi-Private evaluation · medium reasoning	95.50%	Not reported	Closed
44	GPT-5.5 (XHigh)OpenAISemi-Private evaluation · xhigh reasoning	95.00%	Not reported	Unknown
45	GPT-5.5 Pro (XHigh)OpenAISemi-Private evaluation · xhigh reasoning	95.00%	Not reported	Unknown
46	Claude Fable 5.1 (Medium)AnthropicSemi-Private evaluation · medium reasoning	94.50%	Not reported	Unknown
47	Claude Fable 5.2 (Low)AnthropicSemi-Private evaluation · low reasoning	94.50%	Not reported	Unknown
48	DeepSeek V4.1 Flash (Max)DeepSeekSemi-Private evaluation · max reasoning	94.50%	Not reported	Unknown
49	DeepSeek V4.1 Flash 1017 (High)DeepSeekSemi-Private evaluation · high reasoning	94.50%	Not reported	Unknown
50	GPT-5.2 (Refinement)Johan LandSemi-Private evaluation	94.50%	Not reported	Unknown
51	GPT-5.4 Pro (XHigh)OpenAISemi-Private evaluation · xhigh reasoning	94.50%	Not reported	Unknown
52	GPT-5.5 (High)OpenAISemi-Private evaluation · high reasoning	94.50%	Not reported	Unknown
53	Kimi K3 (Max)Moonshot AISemi-Private evaluation · max reasoning	94.50%	Not reported	Open
54	Claude Opus 4.6 (120K, High)AnthropicSemi-Private evaluation	94.00%	Not reported	Unknown
55	GPT-5.6 Terra (XHigh)OpenAISemi-Private evaluation · xhigh reasoning	94.00%	Not reported	Closed
56	GPT-5.4 (XHigh)OpenAISemi-Private evaluation · xhigh reasoning	93.67%	Not reported	Unknown
57	Claude 4.7 (High)AnthropicSemi-Private evaluation · high reasoning	93.50%	Not reported	Unknown
58	DeepSeek V4.1 Flash 1017 (Max)DeepSeekSemi-Private evaluation · max reasoning	93.50%	Not reported	Unknown
59	GPT-6.1 Sol (Low)OpenAISemi-Private evaluation · low reasoning	93.50%	Not reported	Closed
60	Claude Opus 4.6 (120K, Max)AnthropicSemi-Private evaluation	93.00%	Not reported	Unknown
61	GPT-5.4 (High)OpenAISemi-Private evaluation · high reasoning	92.67%	Not reported	Unknown
62	GPT-6 Sol (XHigh)OpenAISemi-Private evaluation · xhigh reasoning	92.67%	Not reported	Closed
63	Claude Fable 5 (Medium)AnthropicSemi-Private evaluation · medium reasoning	92.50%	Not reported	Closed
64	Claude Opus 4.8 (Max)AnthropicSemi-Private evaluation · max reasoning	92.50%	Not reported	Closed
65	Gemini 3.5 Flash (High)GoogleSemi-Private evaluation · high reasoning	92.50%	Not reported	Unknown
66	GPT-5.6 Sol (Medium)OpenAISemi-Private evaluation · medium reasoning	92.50%	Not reported	Closed
67	GPT-5.5 (Medium)OpenAISemi-Private evaluation · medium reasoning	92.17%	Not reported	Unknown
68	Claude 4.7 (Max)AnthropicSemi-Private evaluation · max reasoning	92.00%	Not reported	Unknown
69	Claude Opus 4.6 (120K, Medium)AnthropicSemi-Private evaluation	92.00%	Not reported	Unknown
70	Claude Opus 4.8 (High)AnthropicSemi-Private evaluation · high reasoning	92.00%	Not reported	Closed
71	Claude Sonnet 5.5 (High)AnthropicSemi-Private evaluation · high reasoning	92.00%	Not reported	Unknown
72	GPT-5.6 Terra (High)OpenAISemi-Private evaluation · high reasoning	92.00%	Not reported	Closed
73	Claude Opus 4.8 (Medium)AnthropicSemi-Private evaluation · medium reasoning	91.50%	Not reported	Closed
74	DeepSeek V4.1 Flash 1017 (Low)DeepSeekSemi-Private evaluation · low reasoning	91.50%	Not reported	Unknown
75	Gemini 3.6 Flash (High)GoogleSemi-Private evaluation · high reasoning	91.17%	Not reported	Closed
76	Gemini 3.7 Flash (Medium)GoogleSemi-Private evaluation · medium reasoning	91.17%	Not reported	Closed
77	Claude 4.7 (Low)AnthropicSemi-Private evaluation · low reasoning	91.00%	Not reported	Unknown
78	Claude 4.7 (Medium)AnthropicSemi-Private evaluation · medium reasoning	91.00%	Not reported	Unknown
79	GLM-5.3-Flash (Max)Z.aiSemi-Private evaluation · max reasoning	91.00%	Not reported	Unknown
80	GPT-6 Sol (High)OpenAISemi-Private evaluation · high reasoning	91.00%	Not reported	Closed
81	GPT-5.6 Luna 2026-07-30 (Max)OpenAISemi-Private evaluation · max reasoning	90.67%	Not reported	Unknown
82	Claude Fable 5 (Low)AnthropicSemi-Private evaluation · low reasoning	90.50%	Not reported	Closed
83	DeepSeek V4 Pro 0813 (Low)DeepSeekSemi-Private evaluation · low reasoning	90.50%	Not reported	Open
84	DeepSeek V4.1 Flash (Low)DeepSeekSemi-Private evaluation · low reasoning	90.50%	Not reported	Unknown
85	Gemini 3.8 Flash (Low)GoogleSemi-Private evaluation · low reasoning	90.50%	Not reported	Closed
86	GPT-5.2 Pro (XHigh)OpenAISemi-Private evaluation · xhigh reasoning	90.50%	Not reported	Unknown
87	Grok 4.7 (High)xAISemi-Private evaluation · high reasoning	90.17%	Not reported	Closed
88	Claude Fable 5.1 (Low)AnthropicSemi-Private evaluation · low reasoning	90.00%	Not reported	Unknown
89	DeepSeek V4 Pro 0813 (Max)DeepSeekSemi-Private evaluation · max reasoning	90.00%	Not reported	Open
90	Grok 4.7 (Medium)xAISemi-Private evaluation · medium reasoning	90.00%	Not reported	Closed
91	Grok 4.20 (Reasoning)xAISemi-Private evaluation	89.50%	Not reported	Unknown
92	Grok 4.7 (XHigh)xAISemi-Private evaluation · xhigh reasoning	89.50%	Not reported	Closed
93	Claude Haiku 5.5 (Max)AnthropicSemi-Private evaluation · max reasoning	89.00%	Not reported	Closed
94	DeepSeek V4 Flash 0731 (Max)DeepSeekSemi-Private evaluation · max reasoning	89.00%	Not reported	Open
95	Muse Spark Open (XHigh)Meta AISemi-Private evaluation · xhigh reasoning	89.00%	Not reported	Unknown
96	Claude Opus 5.5 (Low)AnthropicSemi-Private evaluation · low reasoning	88.50%	Not reported	Closed
97	DeepSeek V4.1 Flash (High)DeepSeekSemi-Private evaluation · high reasoning	88.50%	Not reported	Unknown
98	Muse Spark Open (High)Meta AISemi-Private evaluation · high reasoning	88.50%	Not reported	Unknown
99	Claude Opus 4.8 (Low)AnthropicSemi-Private evaluation · low reasoning	88.00%	Not reported	Closed
100	GPT-5.6 Luna (Max)OpenAISemi-Private evaluation · max reasoning	88.00%	Not reported	Closed
101	GPT-5.6 Luna (XHigh)OpenAISemi-Private evaluation · xhigh reasoning	87.67%	Not reported	Closed
102	Claude Haiku 5.5 (XHigh)AnthropicSemi-Private evaluation · xhigh reasoning	87.50%	Not reported	Closed
103	Gemini 3 Deep Think (Preview) ²GoogleSemi-Private evaluation	87.50%	Not reported	Unknown
104	Grok 4.6 (Medium)xAISemi-Private evaluation · medium reasoning	87.50%	Not reported	Closed
105	Qwen3.8-27B (XHigh)AlibabaSemi-Private evaluation · xhigh reasoning	87.50%	Not reported	Unknown
106	Grok 4.5 (Medium)xAISemi-Private evaluation · medium reasoning	87.17%	Not reported	Closed
107	DeepSeek V4 Pro 0813 (High)DeepSeekSemi-Private evaluation · high reasoning	87.17%	Not reported	Open
108	DeepSeek V4 Flash 0731 (High)DeepSeekSemi-Private evaluation · high reasoning	87.00%	Not reported	Open
109	Grok 4.6 (High)xAISemi-Private evaluation · high reasoning	87.00%	Not reported	Closed
110	Grok 4.6 (XHigh)xAISemi-Private evaluation · xhigh reasoning	87.00%	Not reported	Closed
111	GPT-6 Luna (Max)OpenAISemi-Private evaluation · max reasoning	86.67%	Not reported	Closed
112	Kimi K3 (High)Moonshot AISemi-Private evaluation · high reasoning	86.67%	Not reported	Open
113	Claude Sonnet 4.6 (High)AnthropicSemi-Private evaluation · high reasoning	86.50%	Not reported	Unknown
114	GPT-5.2 (XHigh)OpenAISemi-Private evaluation · xhigh reasoning	86.17%	Not reported	Unknown
115	GPT-5.4 (Medium)OpenAISemi-Private evaluation · medium reasoning	86.17%	Not reported	Unknown
116	Claude Opus 4.6 (120K, Low)AnthropicSemi-Private evaluation	86.00%	Not reported	Unknown
117	Claude Sonnet 4.6 (Max)AnthropicSemi-Private evaluation · max reasoning	86.00%	Not reported	Unknown
118	GPT-6 Astra (None)OpenAISemi-Private evaluation · none reasoning	86.00%	Not reported	Closed
119	GPT-5.2 Pro (High)OpenAISemi-Private evaluation · high reasoning	85.67%	Not reported	Unknown
120	Grok 4.5 (High)xAISemi-Private evaluation · high reasoning	85.67%	Not reported	Closed
121	Gemini 4 Carbon (Low)GoogleSemi-Private evaluation · low reasoning	85.50%	Not reported	Closed
122	Gemini 3.7 Flash (Low)GoogleSemi-Private evaluation · low reasoning	85.17%	Not reported	Closed
123	GPT-5.6 Luna 2026-07-30 (XHigh)OpenAISemi-Private evaluation · xhigh reasoning	85.00%	Not reported	Unknown
124	Gemini 3 Flash Preview (High)GoogleSemi-Private evaluation · high reasoning	84.67%	Not reported	Unknown
125	DeepSeek V4 Flash 0731 (Low)DeepSeekSemi-Private evaluation · low reasoning	84.00%	Not reported	Open
126	Inkling Small (XHigh)Thinking Machines LabSemi-Private evaluation · xhigh reasoning	84.00%	Not reported	Open
127	GPT-6 Sol (Medium)OpenAISemi-Private evaluation · medium reasoning	83.67%	Not reported	Closed
128	Gemini 3.6 Flash (Medium)GoogleSemi-Private evaluation · medium reasoning	83.17%	Not reported	Closed
129	Claude Haiku 5.5 (High)AnthropicSemi-Private evaluation · high reasoning	81.50%	Not reported	Closed
130	GPT-5.2 Pro (Medium)OpenAISemi-Private evaluation · medium reasoning	81.17%	Not reported	Unknown
131	Opus 4.5 (Thinking, 64K)AnthropicSemi-Private evaluation	80.00%	Not reported	Unknown
132	Grok 4 (Refinement)J. BermanSemi-Private evaluation	79.60%	Not reported	Unknown
133	InklingThinking MachinesSemi-Private evaluation	79.50%	Not reported	Unknown
134	Grok 4.5 (Low)xAISemi-Private evaluation · low reasoning	79.17%	Not reported	Closed
135	GPT-5.2 (High)OpenAISemi-Private evaluation · high reasoning	78.67%	Not reported	Unknown
136	GPT-5.6 Luna 2026-07-30 (High)OpenAISemi-Private evaluation · high reasoning	78.00%	Not reported	Unknown
137	Inkling Small (High)Thinking Machines LabSemi-Private evaluation · high reasoning	78.00%	Not reported	Open
138	Muse Spark Open (Medium)Meta AISemi-Private evaluation · medium reasoning	77.67%	Not reported	Unknown
139	GLM-5.2Z.aiSemi-Private evaluation	77.00%	Not reported	Unknown
140	GPT-5.6 Terra (Medium)OpenAISemi-Private evaluation · medium reasoning	77.00%	Not reported	Closed
141	Gemini 3.6 Flash (Low)GoogleSemi-Private evaluation · low reasoning	76.50%	Not reported	Closed
142	Gemini 4 Argon (Low)GoogleSemi-Private evaluation · low reasoning	76.50%	Not reported	Closed
143	GPT-5.6 Luna (High)OpenAISemi-Private evaluation · high reasoning	76.50%	Not reported	Closed
144	GPT-5.5 (Low)OpenAISemi-Private evaluation · low reasoning	76.17%	Not reported	Unknown
145	Opus 4.5 (Thinking, 32K)AnthropicSemi-Private evaluation	75.83%	Not reported	Unknown
146	Gemini 3 ProGoogleSemi-Private evaluation	75.00%	Not reported	Unknown
147	Grok 4.6 (Low)xAISemi-Private evaluation · low reasoning	74.83%	Not reported	Closed
148	GPT-5.6 Sol (Low)OpenAISemi-Private evaluation · low reasoning	74.50%	Not reported	Closed
149	Beam 1.0 (XHigh)ReflectionSemi-Private evaluation · xhigh reasoning	74.00%	Not reported	Unknown
150	GPT-6 Luna (XHigh)OpenAISemi-Private evaluation · xhigh reasoning	73.00%	Not reported	Closed
151	GPT-5.1 (Thinking, High)OpenAISemi-Private evaluation	72.83%	Not reported	Unknown
152	GPT-5.2 (Medium)OpenAISemi-Private evaluation · medium reasoning	72.67%	Not reported	Unknown
153	GPT-6 Sol (Low)OpenAISemi-Private evaluation · low reasoning	72.17%	Not reported	Closed
154	Opus 4.5 (Thinking, 16K)AnthropicSemi-Private evaluation	72.00%	Not reported	Unknown
155	GLM-5.3-Flash (High)Z.aiSemi-Private evaluation · high reasoning	71.83%	Not reported	Unknown
156	GPT-6 Luna (High)OpenAISemi-Private evaluation · high reasoning	70.33%	Not reported	Closed
157	GPT-5 ProOpenAISemi-Private evaluation	70.17%	Not reported	Unknown
158	Claude Sonnet 5.5 (Medium)AnthropicSemi-Private evaluation · medium reasoning	70.00%	Not reported	Unknown
159	Qwen3.8-27B (Low)AlibabaSemi-Private evaluation · low reasoning	69.17%	Not reported	Unknown
160	Qwen3.8-27B (Medium)AlibabaSemi-Private evaluation · medium reasoning	68.67%	Not reported	Unknown
161	GPT-5.4 (Low)OpenAISemi-Private evaluation · low reasoning	68.17%	Not reported	Unknown
162	Beam 1.0 (High)ReflectionSemi-Private evaluation · high reasoning	67.67%	Not reported	Unknown
163	Claude Sonnet 5.5 (Low)AnthropicSemi-Private evaluation · low reasoning	67.00%	Not reported	Unknown
164	Inkling Small (Medium)Thinking Machines LabSemi-Private evaluation · medium reasoning	67.00%	Not reported	Open
165	Grok 4 (Thinking)xAISemi-Private evaluation	66.67%	Not reported	Unknown
166	GPT-5 (High)OpenAISemi-Private evaluation · high reasoning	65.67%	Not reported	Unknown
167	Kimi K3 (Low)Moonshot AISemi-Private evaluation · low reasoning	65.67%	Not reported	Open
168	Kimi K2.5Moonshot AISemi-Private evaluation	65.33%	Not reported	Unknown
169	Claude Sonnet 4.5 (Thinking 32K)AnthropicSemi-Private evaluation	63.67%	Not reported	Unknown
170	GPT-5.4 Mini (XHigh)OpenAISemi-Private evaluation · xhigh reasoning	63.67%	Not reported	Unknown
171	Minimax M2.5MinimaxSemi-Private evaluation	63.67%	Not reported	Unknown
172	Grok 4.7 (Low)xAISemi-Private evaluation · low reasoning	63.33%	Not reported	Closed
173	GPT-6 Luna (Medium)OpenAISemi-Private evaluation · medium reasoning	61.00%	Not reported	Closed
174	o3 (High)OpenAISemi-Private evaluation · high reasoning	60.83%	Not reported	Unknown
175	GPT-5.6 Terra (Low)OpenAISemi-Private evaluation · low reasoning	60.17%	Not reported	Closed
176	o3-Pro (High)OpenAISemi-Private evaluation · high reasoning	59.33%	Not reported	Unknown
177	o4-mini (High)OpenAISemi-Private evaluation · high reasoning	58.67%	Not reported	Unknown
178	Opus 4.5 (Thinking, 8K)AnthropicSemi-Private evaluation	58.67%	Not reported	Unknown
179	Muse Spark Open (Low)Meta AISemi-Private evaluation · low reasoning	58.50%	Not reported	Unknown
180	GPT-5.4 Mini (High)OpenAISemi-Private evaluation · high reasoning	58.00%	Not reported	Unknown
181	Gemini 3 Flash Preview (Medium)GoogleSemi-Private evaluation · medium reasoning	57.67%	Not reported	Unknown
182	GPT-5.1 (Thinking, Medium)OpenAISemi-Private evaluation	57.67%	Not reported	Unknown
183	Claude Haiku 5.5 (Medium)AnthropicSemi-Private evaluation · medium reasoning	57.33%	Not reported	Closed
184	Deepseek V3.2DeepSeekSemi-Private evaluation	57.00%	Not reported	Unknown
185	o3-Pro (Medium)OpenAISemi-Private evaluation · medium reasoning	57.00%	Not reported	Unknown
186	GPT-5.6 Luna (Medium)OpenAISemi-Private evaluation · medium reasoning	56.50%	Not reported	Closed
187	GPT-5 (Medium)OpenAISemi-Private evaluation · medium reasoning	56.17%	Not reported	Unknown
188	ARChitectsARC Prize 2024Semi-Private evaluation	56.00%	Not reported	Unknown
189	GPT-5.2 (Low)OpenAISemi-Private evaluation · low reasoning	55.67%	Not reported	Unknown
190	Beam 1.0 (Medium)ReflectionSemi-Private evaluation · medium reasoning	55.67%	Not reported	Unknown
191	GPT-5 Mini (High)OpenAISemi-Private evaluation · high reasoning	54.33%	Not reported	Unknown
192	GPT-5.6 Luna 2026-07-30 (Medium)OpenAISemi-Private evaluation · medium reasoning	54.00%	Not reported	Unknown
193	o3 (Medium)OpenAISemi-Private evaluation · medium reasoning	53.83%	Not reported	Unknown
194	Gemini 3.5 Flash-Lite (High)GoogleSemi-Private evaluation · high reasoning	53.50%	Not reported	Unknown
195	Inkling Small (Low)Thinking Machines LabSemi-Private evaluation · low reasoning	52.50%	Not reported	Open
196	GPT-5.4 Nano (XHigh)OpenAISemi-Private evaluation · xhigh reasoning	51.50%	Not reported	Unknown
197	Claude Haiku 5.5 (Low)AnthropicSemi-Private evaluation · low reasoning	50.33%	Not reported	Closed
198	Gemini 3.5 Flash (Minimal)GoogleSemi-Private evaluation · minimal reasoning	48.83%	Not reported	Unknown
199	Grok 4 (Fast Reasoning)xAISemi-Private evaluation	48.50%	Not reported	Unknown
200	Inkling Small (Minimal)Thinking Machines LabSemi-Private evaluation · minimal reasoning	48.50%	Not reported	Open
201	Claude Sonnet 4.5 (Thinking 16K)AnthropicSemi-Private evaluation	48.33%	Not reported	Unknown
202	Claude Haiku 4.5 (Thinking 32K)AnthropicSemi-Private evaluation	47.67%	Not reported	Unknown
203	GLM-5.3-Flash (Low)Z.aiSemi-Private evaluation · low reasoning	47.00%	Not reported	Unknown
204	Claude Sonnet 4.5 (Thinking 8K)AnthropicSemi-Private evaluation	46.50%	Not reported	Unknown
205	Beam 1.0 (Low)ReflectionSemi-Private evaluation · low reasoning	45.00%	Not reported	Unknown
206	GLM-5Z.aiSemi-Private evaluation	44.67%	Not reported	Unknown
207	o3-Pro (Low)OpenAISemi-Private evaluation · low reasoning	44.33%	Not reported	Unknown
208	GPT-5 (Low)OpenAISemi-Private evaluation · low reasoning	44.00%	Not reported	Unknown
209	o4-mini (Medium)OpenAISemi-Private evaluation · medium reasoning	41.83%	Not reported	Unknown
210	o3 (Low)OpenAISemi-Private evaluation · low reasoning	41.50%	Not reported	Unknown
211	Gemini 2.5 Pro (Thinking 16K)GoogleSemi-Private evaluation	41.00%	Not reported	Unknown
212	GPT-5.4 Mini (Medium)OpenAISemi-Private evaluation · medium reasoning	40.83%	Not reported	Unknown
213	Claude Sonnet 4 (Thinking 16K)AnthropicSemi-Private evaluation	40.00%	Not reported	Unknown
214	Opus 4.5 (Thinking, None)AnthropicSemi-Private evaluation	40.00%	Not reported	Unknown
215	Tiny Recursion Model (TRM)BespokeSemi-Private evaluation	40.00%	Not reported	Unknown
216	GPT-5.6 Luna 2026-07-30 (Low)OpenAISemi-Private evaluation · low reasoning	39.50%	Not reported	Unknown
217	GPT-5.4 Nano (High)OpenAISemi-Private evaluation · high reasoning	38.17%	Not reported	Unknown
218	GPT-6 Luna (Low)OpenAISemi-Private evaluation · low reasoning	37.67%	Not reported	Closed
219	Claude Haiku 4.5 (Thinking 16K)AnthropicSemi-Private evaluation	37.33%	Not reported	Unknown
220	GPT-5 Mini (Medium)OpenAISemi-Private evaluation · medium reasoning	37.33%	Not reported	Unknown
221	Gemini 2.5 Pro (Thinking 32K)GoogleSemi-Private evaluation	37.00%	Not reported	Unknown
222	Claude Opus 4 (Thinking 16K)AnthropicSemi-Private evaluation	35.67%	Not reported	Unknown
223	Gemini 3.6 Flash (Minimal)GoogleSemi-Private evaluation · minimal reasoning	34.50%	Not reported	Closed
224	o3-mini (High)OpenAISemi-Private evaluation · high reasoning	34.50%	Not reported	Unknown
225	GPT-5.6 Luna (Low)OpenAISemi-Private evaluation · low reasoning	34.17%	Not reported	Closed
226	Qwen3.8-27B (None)AlibabaSemi-Private evaluation · none reasoning	34.00%	Not reported	Unknown
227	Gemini 2.5 Flash (Preview)GoogleSemi-Private evaluation	33.33%	Not reported	Unknown
228	Gemini 2.5 Flash (Preview) (Thinking 16K)GoogleSemi-Private evaluation	33.33%	Not reported	Unknown
229	GPT-5.1 (Thinking, Low)OpenAISemi-Private evaluation	33.17%	Not reported	Unknown
230	GPT-5.4 Nano (Medium)OpenAISemi-Private evaluation · medium reasoning	33.00%	Not reported	Unknown
231	Gemini 3.5 Flash-Lite (Medium)GoogleSemi-Private evaluation · medium reasoning	32.33%	Not reported	Unknown
232	Gemini 2.5 Flash (Preview) (Thinking 24K)GoogleSemi-Private evaluation	32.33%	Not reported	Unknown
233	Claude Sonnet 4.5 (Thinking 1K)AnthropicSemi-Private evaluation	31.00%	Not reported	Unknown
234	Claude Opus 4 (Thinking 8K)AnthropicSemi-Private evaluation	30.67%	Not reported	Unknown
235	Gemini 2.5 Pro (Thinking 8K)GoogleSemi-Private evaluation	29.50%	Not reported	Unknown
236	GPT-6 Sol (None)OpenAISemi-Private evaluation · none reasoning	29.33%	Not reported	Closed
237	Claude Sonnet 4 (Thinking 8K)AnthropicSemi-Private evaluation	29.00%	Not reported	Unknown
238	Gemini 3 Flash Preview (Low)GoogleSemi-Private evaluation · low reasoning	29.00%	Not reported	Unknown
239	Claude 3.7 (16K)AnthropicSemi-Private evaluation	28.60%	Not reported	Unknown
240	Claude Sonnet 4 (Thinking 1K)AnthropicSemi-Private evaluation	28.00%	Not reported	Unknown
241	Codex Mini (Latest)OpenAISemi-Private evaluation	27.33%	Not reported	Unknown
242	Claude Opus 4 (Thinking 1K)AnthropicSemi-Private evaluation	27.00%	Not reported	Unknown
243	GPT-5 Mini (Low)OpenAISemi-Private evaluation · low reasoning	26.33%	Not reported	Unknown
244	Gemini 2.5 Flash (Preview) (Thinking 8K)GoogleSemi-Private evaluation	25.83%	Not reported	Unknown
245	Claude Haiku 4.5 (Thinking 8K)AnthropicSemi-Private evaluation	25.50%	Not reported	Unknown
246	Claude Sonnet 4.5AnthropicSemi-Private evaluation	25.50%	Not reported	Unknown
247	Claude Sonnet 4AnthropicSemi-Private evaluation	23.83%	Not reported	Unknown
248	Nemotron 3.5 Super VL (High)NVIDIASemi-Private evaluation · high reasoning	22.67%	Not reported	Unknown
249	Claude Opus 4AnthropicSemi-Private evaluation	22.50%	Not reported	Unknown
250	o3-mini (Medium)OpenAISemi-Private evaluation · medium reasoning	22.33%	Not reported	Unknown
251	Gemini 3 Flash Preview (Minimal)GoogleSemi-Private evaluation · minimal reasoning	21.50%	Not reported	Unknown
252	o4-mini (Low)OpenAISemi-Private evaluation · low reasoning	21.33%	Not reported	Unknown
253	Deepseek R1 (05/28)DeepSeekSemi-Private evaluation	21.21%	Not reported	Unknown
254	Claude 3.7 (8K)AnthropicSemi-Private evaluation	21.20%	Not reported	Unknown
255	GPT-5 Nano (Medium)OpenAISemi-Private evaluation · medium reasoning	20.71%	Not reported	Unknown
256	Nemotron 3.5 Super VL (Medium)NVIDIASemi-Private evaluation · medium reasoning	20.17%	Not reported	Unknown
257	GPT-5.4 Nano (Low)OpenAISemi-Private evaluation · low reasoning	18.33%	Not reported	Unknown
258	Gemini 3.5 Flash-Lite (Low)GoogleSemi-Private evaluation · low reasoning	17.00%	Not reported	Unknown
259	IcecuberARC Prize 2024Semi-Private evaluation	17.00%	Not reported	Unknown
260	Claude Haiku 4.5 (Thinking 1K)AnthropicSemi-Private evaluation	16.83%	Not reported	Unknown
261	GPT-5 Nano (High)OpenAISemi-Private evaluation · high reasoning	16.67%	Not reported	Unknown
262	Grok 3 Mini (Low)xAISemi-Private evaluation · low reasoning	16.50%	Not reported	Unknown
263	Gemini 2.5 Flash (Preview) (Thinking 1K)GoogleSemi-Private evaluation	16.00%	Not reported	Unknown
264	Gemini 2.5 Pro (Thinking 1K)GoogleSemi-Private evaluation	16.00%	Not reported	Unknown
265	Deepseek R1DeepSeekSemi-Private evaluation	15.80%	Not reported	Unknown
266	o3-mini (Low)OpenAISemi-Private evaluation · low reasoning	14.50%	Not reported	Unknown
267	Claude Haiku 4.5AnthropicSemi-Private evaluation	14.33%	Not reported	Unknown
268	o1-miniOpenAISemi-Private evaluation	14.00%	Not reported	Unknown
269	Claude 3.7AnthropicSemi-Private evaluation	13.60%	Not reported	Unknown
270	DeepSeek V4 Pro 0813 (None)DeepSeekSemi-Private evaluation · none reasoning	13.00%	Not reported	Open
271	GPT-5.4 Mini (Low)OpenAISemi-Private evaluation · low reasoning	13.00%	Not reported	Unknown
272	GPT-5.2OpenAISemi-Private evaluation	12.33%	Not reported	Unknown
273	DeepSeek V4 Flash 0731 (None)DeepSeekSemi-Private evaluation · none reasoning	11.83%	Not reported	Open
274	Claude 3.7 (1K)AnthropicSemi-Private evaluation	11.60%	Not reported	Unknown
275	Qwen3-235b-a22b Instruct (25/07)AlibabaSemi-Private evaluation	11.00%	Not reported	Unknown
276	GPT-4.5OpenAISemi-Private evaluation	10.30%	Not reported	Unknown
277	GPT-6 Luna (None)OpenAISemi-Private evaluation · none reasoning	8.83%	Not reported	Closed
278	DeepSeek V4.1 Flash (None)DeepSeekSemi-Private evaluation · none reasoning	7.83%	Not reported	Unknown
279	Gemini 3.5 Flash-Lite (Minimal)GoogleSemi-Private evaluation · minimal reasoning	7.50%	Not reported	Unknown
280	Magistral Medium (Thinking)MistralSemi-Private evaluation	6.12%	Not reported	Unknown
281	GPT-5 (Minimal)OpenAISemi-Private evaluation · minimal reasoning	6.00%	Not reported	Unknown
282	Magistral MediumMistralSemi-Private evaluation	5.91%	Not reported	Unknown
283	GPT-5.1 (Thinking, None)OpenAISemi-Private evaluation	5.83%	Not reported	Unknown
284	DeepSeek V4.1 Flash 1017 (None)DeepSeekSemi-Private evaluation · none reasoning	5.67%	Not reported	Unknown
285	GPT-4.1OpenAISemi-Private evaluation	5.50%	Not reported	Unknown
286	Grok 3xAISemi-Private evaluation	5.50%	Not reported	Unknown
287	Inkling Small (None)Thinking Machines LabSemi-Private evaluation · none reasoning	5.50%	Not reported	Open
288	GPT-5 Mini (Minimal)OpenAISemi-Private evaluation · minimal reasoning	5.33%	Not reported	Unknown
289	Magistral SmallMistralSemi-Private evaluation	5.00%	Not reported	Unknown
290	GPT-4oOpenAISemi-Private evaluation	4.50%	Not reported	Unknown
291	Llama 4 MaverickMetaSemi-Private evaluation	4.38%	Not reported	Unknown
292	GPT-5 Nano (Low)OpenAISemi-Private evaluation · low reasoning	4.04%	Not reported	Unknown
293	GPT-4.1-MiniOpenAISemi-Private evaluation	3.50%	Not reported	Unknown
294	GPT-5.6 Luna 2026-07-30 (None)OpenAISemi-Private evaluation · none reasoning	3.33%	Not reported	Unknown
295	Nemotron 3.5 Super VL (None)NVIDIASemi-Private evaluation · none reasoning	2.00%	Not reported	Unknown
296	GPT-5 Nano (Minimal)OpenAISemi-Private evaluation · minimal reasoning	1.50%	Not reported	Unknown
297	Llama 4 ScoutMetaSemi-Private evaluation	0.50%	Not reported	Unknown
298	GPT-4.1-NanoOpenAISemi-Private evaluation	0.00%	Not reported	Unknown

Semi-Private configurations

These are ARC Prize's published Semi-Private results. Every published reasoning-effort configuration stays separate; the table does not combine efforts or evaluation splits.

ARC-AGI-1 and ARC-AGI-2 use different task sets and remain separate from the interactive ARC-AGI-3 track.

Model pages use the highest declared published reasoning effort within each harness, rather than selecting whichever effort scored highest. Hidden or unpublished incomplete runs are not assigned a score.

Snapshot

ARC Prize leaderboard Original aggregate results

The published ARC-AGI-1 snapshot places Claude Fable 5 (Max) first at 98.50%. The third row is 0.00 points behind. The broader top-10 range is 0.00 points, so many of the published results sit in a relatively narrow band.

298 configurations have been evaluated on ARC-AGI-1. The benchmark falls in the Reasoning category. ARC-AGI-1 is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.

About ARC-AGI-1

Year

2026

Tasks

Semi-private ARC-AGI-1 evaluation set

Format

Verified accuracy

Difficulty

Abstract visual reasoning

ARC Prize publishes verified semi-private results by model and reasoning effort. The model evidence note identifies the effort for each sourced row.

ARC Prize leaderboard Public benchmark source

Freshness and provenance

Version

ARC-AGI-1 2026

Refresh cadence

Quarterly

Staleness state

Current

Question availability

Public benchmark set

CurrentDisplay only

BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.

Questions

What does ARC-AGI-1 measure?

ARC Prize fluid-intelligence benchmark using novel visual grid transformations.

Which model leads the published ARC-AGI-1 snapshot?

Claude Fable 5 (Max) currently leads the published ARC-AGI-1 snapshot with 98.50% semi-private score. BenchLM shows this benchmark for display only and does not use it in overall rankings.

How many models are evaluated on ARC-AGI-1?

The Captured 2026-11-10 snapshot contains 298 AI models.

Last updated: Captured 2026-11-10 · mirrored from the public benchmark leaderboard

Know when it’s worth switching models

The model to choose, the cheaper alternative, and the release we would wait on.

Read a sample issue

Join 7,000+ subscribers.

One email each week. Unsubscribe anytime.