SShogo Bench

ARC-AGI

The original ARC abstraction and reasoning corpus.

Agent Index weight
0%
shown for context only
Models tested
95
from the benchmark’s leaderboard and reports
Normalised for the index
0 → 0, 1 → 100
Official site

Results

#ModelScoreNormalisedHarnessCost/taskDate
1Claude Fable 5 Supported
Anthropic
98.5%source——
2Claude Fable 5.2 Supported
Anthropic
98.5%source——
3Claude Opus 5.5 Supported
Anthropic
98.5%source——
4Gemini 3.8 Flash Supported
Google
98.5%source——
5Gemini 4 Argon Supported
Google
98.5%source——
6Gemini 4 Carbon Supported
Google
98.5%source——
7GPT-6.1 Sol Estimated
OpenAI
98.5%source——
8GPT-6 Astra Supported
OpenAI
98.5%source——
9Gemini 3.1 Pro Preview Supported
Google
98.0%source$0.52—
10Claude Fable 5.1 Supported
Anthropic
97.5%source——
11Claude Opus 5 Supported
Anthropic
97.5%source——
12GPT-5.6 Sol Supported
OpenAI
97.5%source$0.40—
13Claude Sonnet 5.5 Estimated
Anthropic
96.5%source——
14GPT-5.5 Pro
OpenAI
96.5%source$4.53—
15GPT-5.6 Terra Estimated
OpenAI
96.5%source$0.55—
16Inkling 2 Small Estimated
thinkingmachines
96.5%source——
17Gemini 3 Deep Think
Google
96.0%source——
18Gemini 3.7 Flash Estimated
Google
95.5%source——
19GPT-6 Sol Estimated
OpenAI
95.5%source——
20GPT-5.5 Supported
OpenAI
95.0%source——
21DeepSeek V4.1 Flash Estimated
DeepSeek
94.5%source——
22DeepSeek V4.1 Flash 1017 Estimated
DeepSeek
94.5%source——
23GPT-5.4 Pro
OpenAI
94.5%source——
24Kimi K3 Supported
Moonshot AI
94.5%source$0.77—
25Claude Opus 4.6 Supported
Anthropic
94.0%source——
26GPT-5.4 Supported
OpenAI
93.7%source——
27Claude Opus 4.7 Supported
Anthropic
93.5%source$1.41—
28Claude Opus 4.8 Supported
Anthropic
92.5%source$2.33—
29Gemini 3.5 Flash Supported
Google
92.5%source$0.43—
30Gemini 3.6 Flash Estimated
Google
91.2%source——
31GLM-5.3-Flash Supported
Z.ai (Zhipu)
91.0%source——
32DeepSeek V4 Pro 0813
DeepSeek
90.5%source——
33GPT-5.2 Pro
OpenAI
90.5%source$11.65—
34Grok 4.7 Estimated
xAI
90.2%source——
35Grok 4.20 (Non-Reasoning) Estimated
xAI
89.5%source$0.35—
36Claude Haiku 5.5 Estimated
Anthropic
89.0%source——
37DeepSeek V4 Flash 0731
DeepSeek
89.0%source——
38Muse Spark Open
Meta
89.0%source——
39GPT-5.6 Luna Estimated
OpenAI
88.0%source$0.32—
40Grok 4.6 Supported
xAI
87.5%source——
41Qwen3.8 27B
Alibaba (Qwen)
87.5%source——
42Grok 4.5 Estimated
xAI
87.2%source$0.31—
43GPT-6 Luna
OpenAI
86.7%source——
44Claude Sonnet 4.6 Supported
Anthropic
86.5%source$1.19—
45GPT-5.2 Supported
OpenAI
86.2%source$0.96—
46Gemini 3 Flash Preview Supported
Google
84.7%source$0.17—
47Inkling Small
thinkingmachines
84.0%source——
48Claude Opus 4.5 (latest) Supported
Anthropic
80.0%source$1.47—
49Inkling Estimated
thinkingmachines
79.5%source$0.30—
50GLM-5.2 Supported
Z.ai (Zhipu)
77.0%source$0.19—
51Gemini 3 Pro Preview Supported
Google
75.0%source$0.49—
52Beam 1.0
Reflection
74.0%source——
53GPT-5.1 Supported
OpenAI
72.8%source——
54GPT-5 Pro
OpenAI
70.2%source——
55Grok 4 Supported
xAI
66.7%source$2.17—
56GPT-5 Supported
OpenAI
65.7%source$0.51—
57Kimi K2.5 Supported
Moonshot AI
65.3%source$0.15—
58Claude Sonnet 4.5 (latest) Supported
Anthropic
63.7%source$0.52—
59GPT-5.4 mini
OpenAI
63.7%source$0.51—
60MiniMax-M2.5 Estimated
MiniMax
63.7%source$0.07—
61o3 Supported
OpenAI
60.8%source$0.50—
62o3-pro
OpenAI
59.3%source$4.16—
63o4-mini Estimated
OpenAI
58.7%source$0.41—
64DeepSeek V3.2 Supported
DeepSeek
57.0%source$0.08—
65GPT-5 Mini Estimated
OpenAI
54.3%source$0.12—
66Gemini 3.5 Flash Lite
Google
53.5%source——
67GPT-5.4 nano
OpenAI
51.5%source$0.13—
68Grok 4 Fast
xAI
48.5%source$0.03—
69Claude Haiku 4.5 Estimated
Anthropic
47.7%source$0.26—
70GLM-5 Supported
Z.ai (Zhipu)
44.7%source$0.17—
71Gemini 2.5 Pro Supported
Google
41.0%source$0.48—
72Claude Sonnet 4 Supported
Anthropic
40.0%source$0.37—
73tiny-recursion-model
Unknown
40.0%source$1.76—
74Claude-Opus-4 Supported
Anthropic
35.7%source$1.25—
75o3-mini Estimated
OpenAI
34.5%source$0.40—
76Gemini 2.5 Flash Estimated
Google
33.3%source$0.21—
77o1 Estimated
OpenAI
30.7%source——
78Claude 3.7 Sonnet Supported
Anthropic
28.6%source$0.33—
79codex-mini-2025-05-16
Unknown
27.3%source$0.16—
80o1-pro
OpenAI
23.3%source——
81DeepSeek-R1 Estimated
DeepSeek
21.2%source$0.05—
82GPT-5 Nano
OpenAI
20.7%source$0.01—
83Grok-3 mini
xAI
16.5%source——
84o1-mini
OpenAI
14.0%source——
85Qwen3 235B A22B Instruct 2507
Alibaba (Qwen)
11.0%source$0.00—
86GPT-4.5
OpenAI
10.3%source$0.29—
87Magistral Medium 1.1
Mistral
6.1%source$0.12—
88GPT-4.1
OpenAI
5.5%source$0.04—
89Grok 3
xAI
5.5%source$0.09—
90Magistral Small 1.0
Mistral
5.0%source$0.04—
91GPT-4o (2024-08-06) Supported
OpenAI
4.5%source$0.05—
92Llama 4 Maverick
Meta
4.4%source$0.01—
93GPT-4.1 mini
OpenAI
3.5%source$0.01—
94Llama 4 Scout
Meta
0.5%source$0.00—
95GPT-4.1 nano
OpenAI
0.0%source$0.00—