VECTOR WIREAI INTELLIGENCE
⌘K
NVDA$1,847+3.2%MSFT$512+1.1%GOOGL$199-0.4%META$728+2.7%AMD$184-1.2%TSM$212+0.6%PLTR$98+4.1%AI IDX4,821+1.9%
PKT
SEEDRefresh Models —Deals —Regulatory —Sources —

critpt

100 models tracked

CritPt — Physics research-level reasoning benchmark; broad multi-source aggregate, base tools axis

#ModelVendorBest scoreRunsLast seen
1GPT-5.6 SolOpenAI32.372026-07-27
2GPT-5.5 ProOpenAI30.612026-05-02
3GPT-5.6 TerraOpenAI3062026-07-10
4GPT-5.4 ProOpenAI3012026-05-02
5Opus 5Anthropic29.152026-07-25
6Claude Mythos 5Anthropic28.612026-06-10
7Claude Fable 5Anthropic28.622026-07-27
8GPT-5.5OpenAI27.192026-07-27
9Gemini 3 Deep ThinkGoogle25.712026-05-02
10Kimi K3Moonshot23.442026-07-31
11GPT-5.4OpenAI23.432026-05-02
12GLM-5.2 Full Open SourceZ.ai20.982026-08-23
13Claude Opus 4.8Anthropic20.942026-07-27
14GPT-5.6 LunaOpenAI20.672026-07-30
15Qwen3.8 2.4T A95BAlibaba2012026-08-14
16Qwen3.8 Max PreviewAlibaba2012026-08-06
17Grok 4.6SpaceXAI19.742026-08-22
18GLM-5.3Z.ai19.112026-08-19
19Gemini 3.1 ProGoogle17.762026-07-13
20Muse Spark 1.2Meta17.712026-08-06
21GPT-5.3 CodexOpenAI16.912026-05-02
22Claude Sonnet 5Anthropic16.922026-07-02
23DeepSeek-V4-FlashDeepSeek16.672026-08-01
24Grok4.5SpaceXAI15.412026-07-09
25Muse Spark 1.1Meta15.112026-07-11
26Gemini 3.7 FlashGoogle14.332026-08-14
27Qwen3.7 MaxAlibaba13.442026-08-23
28Gemini 3.5 FlashGoogle13.142026-07-07
29DeepSeek-V4-ProDeepSeek1362026-07-30
30Claude Opus 4.6Anthropic12.622026-06-01
31Claude Opus 4.7Anthropic1222026-06-01
32GPT-5.2OpenAI11.632026-05-02
33Muse SparkMeta11.322026-07-07
34Agnes 2.5 Pro AlphaSapiens AI10.912026-07-25
35Gemini 3.6 FlashGoogle10.612026-07-22
36GPT-5.4 miniOpenAI1032026-06-01
37Kimi K2.7 CodeMoonshot1022026-07-07
38GPT-5.4 nanoOpenAI9.332026-06-01
39Grok Build 0.1SpaceXAI9.122026-07-07
40Qwen3.7 Plus PreviewAlibaba9.132026-08-23
41Gemini 3 ProGoogle9.122026-05-02
42A.X-K2SK Telecom8.912026-08-13
43GPT-5.2 CodexOpenAI8.712026-05-02
44Nex-N2-ProNex AGI8.622026-07-07
45Gemini 3 Flash PreviewGoogle8.622026-05-02
46Inkling-SmallThinking Machines8.342026-08-23
47Kimi K2.6Moonshot842026-07-30
48Grok 4.3SpaceXAI852026-05-23
49DeepSeek-V3.2-SpecialeDeepSeek7.422026-07-07
50Grok 4.20 0309 V2SpaceXAI6.632026-06-18
51Motif-3Motif Technologies6.612026-08-13
52Motif-3-BetaMotif Technologies612026-07-21
53Grok 4.20SpaceXAI612026-05-02
54Solar Open2 250BUpstage5.712026-08-13
55GPT-5OpenAI5.762026-07-07
56GPT-5.1 CodexOpenAI5.712026-05-02
57Solar Pro4Upstage5.412026-08-13
58InklingThinking Machines5.422026-07-30
59Qwen3.8 27BAlibaba5.442026-08-22
60GPT-5 CodexOpenAI5.112026-05-02
61GPT-5.1OpenAI4.922026-06-20
62Hy3-previewTencent4.922026-05-02
63GLM-5.1Z.ai4.642026-07-13
64Claude Opus 4.5Anthropic4.622026-05-06
65MiMo V2 FlashXiaomi4.342026-07-07
66MiMo V2.5 ProXiaomi452026-07-30
67MiniMax M3MiniMax3.752026-07-30
68Qwen3.6 Max PreviewAlibaba3.712026-05-02
69Nemotron 3 Super 120B A12BNVIDIA3.112026-06-01
70Nemotron 3 Ultra 550B A55BNVIDIA3.112026-06-06
71Kimi K2.5Moonshot3.132026-07-30
72Claude Sonnet 4.6Anthropic3.132026-05-02
73DeepSeek-V3.2DeepSeek2.922026-05-02
74Qwen3.6 PlusAlibaba2.922026-07-07
75Grok 4.1 FastSpaceXAI2.922026-06-23
76Grok 4 FastSpaceXAI2.922026-05-02
77Kimi K2 (Reasoning)Moonshot2.622026-07-07
78Gemini 2.5 ProGoogle2.622026-07-07
79LongCat-2.0Meituan2.612026-07-20
80GPT-5.5 InstantOpenAI2.642026-07-07
81Muse GlimmerMeta2.612026-08-11
82Step 3.5 FlashStepFun2.542026-07-07
83Step 3.7 FlashStepFun2.322026-07-07
84Grok 4SpaceXAI222026-07-07
85GLM-5Z.ai222026-05-02
86DeepSeek-V3.1DeepSeek222026-06-01
87DeepSeek-V3.1-TerminusDeepSeek1.722026-06-01
88Ling-3.0-flashInclusionAI1.712026-08-06
89GLM-4.7Z.ai1.722026-06-01
90Qwen3.5 397B A17BAlibaba1.742026-07-30
91DeepSeek-V3.2-ExpDeepSeek1.422026-06-01
92GPT-5 miniOpenAI1.452026-07-07
93gpt-oss-20bOpenAI1.422026-05-02
94Gemma 4 31BGoogle1.422026-06-01
95DeepSeek-R1 0528 (May '25)DeepSeek1.422026-07-07
96ERNIE 5.0 Thinking PreviewBaidu1.412026-05-02
97Gemini 2.5 FlashGoogle1.422026-05-02
98Qwen3.6 27BAlibaba1.122026-05-02
99o3OpenAI1.122026-07-07
100GLM-4.6Z.ai1.122026-06-19

Best tracked score per model (default configuration; source-attributed and verification-tiered). Open a model for its full benchmark surface, provenance and pricing.