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 —

OmniScience Accuracy

100 models tracked

OmniScience (Artificial Analysis scientific-knowledge benchmark) — AA-sourced scientific-knowledge accuracy; distinct metric from Non-Hallucination.

#ModelVendorBest scoreRunsLast seen
1Claude Fable 5Anthropic65.312026-06-10
2Opus 5Anthropic60.952026-07-25
3GPT-5.6 SolOpenAI59.462026-07-10
4GPT-5.5OpenAI5852026-06-01
5Gemini 3 ProGoogle55.822026-06-01
6Gemini 3.7 FlashGoogle55.332026-08-14
7Gemini 3.1 ProGoogle55.322026-07-07
8Gemini 3 Flash PreviewGoogle53.422026-06-01
9GPT-5.3 CodexOpenAI52.912026-06-01
10Muse Spark 1.1Meta5212026-07-11
11Grok4.5SpaceXAI51.512026-07-09
12Grok Build 0.1SpaceXAI51.522026-07-07
13Gemini 3.5 FlashGoogle51.442026-07-07
14GPT-5.4OpenAI50.932026-06-01
15Gemini 3.6 FlashGoogle5012026-07-22
16Muse SparkMeta49.622026-07-07
17Claude Opus 4.7Anthropic48.922026-06-01
18Claude Opus 4.8Anthropic48.812026-06-01
19Grok 4.6SpaceXAI48.242026-08-22
20Kimi K3Moonshot47.632026-07-31
21Claude Opus 4.6Anthropic4722026-06-01
22GPT-5.6 TerraOpenAI46.862026-07-10
23Claude Opus 4.5Anthropic46.622026-06-01
24GPT-5.5 InstantOpenAI46.342026-07-07
25Muse Spark 1.2Meta45.412026-08-06
26GPT-5.2OpenAI44.332026-06-01
27DeepSeek-V4-ProDeepSeek4332026-06-29
28GPT-5.6 LunaOpenAI42.762026-07-10
29InklingThinking Machines41.512026-07-16
30Grok 4SpaceXAI41.422026-07-07
31GPT-5.2 CodexOpenAI41.112026-06-01
32Claude Sonnet 4.6Anthropic40.932026-06-01
33DeepSeek-V4-FlashDeepSeek40.452026-08-01
34GPT-5OpenAI40.362026-07-07
35Claude Sonnet 5Anthropic4022026-07-02
36GPT-5.1 CodexOpenAI39.912026-06-01
37Kimi K2.7 CodeMoonshot39.622026-07-07
38Gemini 2.5 ProGoogle3922026-07-07
39GPT-5 CodexOpenAI38.712026-06-01
40o3OpenAI38.622026-07-07
41DeepSeek-V3.2-SpecialeDeepSeek38.422026-07-07
42Qwen3.6 Max PreviewAlibaba37.912026-06-01
43GPT-5.1OpenAI37.722026-06-20
44GPT-5.4 miniOpenAI37.532026-06-01
45Gemini 3.1 Flash Lite PreviewGoogle36.422026-07-07
46Kimi K2.5Moonshot35.222026-06-01
47Nex-N2-ProNex AGI34.822026-07-07
48Grok 4.3SpaceXAI34.842026-06-01
49o1OpenAI34.722026-07-07
50GLM-5.3Z.ai33.912026-08-19
51Agnes 2.5 Pro AlphaSapiens AI33.512026-07-25
52Inkling-SmallThinking Machines33.212026-07-31
53DeepSeek-V3.2DeepSeek3322026-06-01
54Claude Sonnet 4.5Anthropic32.922026-06-01
55Kimi K2.6Moonshot32.832026-07-07
56Hy3-previewTencent3222026-06-01
57Qwen3.8 Max PreviewAlibaba31.912026-08-06
58Qwen3.8 2.4T A95BAlibaba31.312026-08-14
59Qwen3.7 MaxAlibaba31.112026-06-01
60DeepSeek-R1 0528 (May '25)DeepSeek30.922026-07-07
61Kimi K2 (Reasoning)Moonshot30.922026-07-07
62Qwen3.5 397B A17BAlibaba30.822026-06-01
63DeepSeek-R1DeepSeek30.722026-07-07
64Motif-3Motif Technologies30.112026-08-13
65LongCat-2.0Meituan29.612026-07-20
66Gemini 3.5 Flash LiteGoogle29.512026-07-22
67GLM 5V TurboZ.ai29.312026-06-01
68GLM-4.7Z.ai29.322026-06-01
69GLM-5-TurboZ.ai2922026-07-07
70DeepSeek-V3.1DeepSeek2922026-06-01
71Grok 3SpaceXAI28.922026-07-07
72Grok 4.20SpaceXAI28.912026-06-01
73Gemini 2.5 Flash Preview 09-2025Google28.322026-06-01
74Claude 3.7 SonnetAnthropic28.222026-06-01
75GPT-4.1OpenAI27.822026-07-07
76DeepSeek-V3.1-TerminusDeepSeek27.722026-06-01
77DeepSeek-V3.2-ExpDeepSeek27.622026-06-01
78Grok 4.20 0309 V2SpaceXAI27.532026-06-18
79Qwen3 Max (Reasoning)Alibaba27.422026-07-07
80Kimi K2 (Non-Reasoning)Moonshot27.422026-06-19
81MiMo V2.5 ProXiaomi27.232026-07-07
82Muse GlimmerMeta2712026-08-11
83GLM-4.6Z.ai26.922026-06-19
84MiMo V2 ProXiaomi26.822026-07-07
85MiniMax M2.7MiniMax26.822026-07-07
86Qwen3.6 PlusAlibaba26.422026-07-07
87GLM-5Z.ai26.322026-06-01
88MiniMax M2.5MiniMax26.222026-07-07
89Gemini 2.5 FlashGoogle26.122026-06-01
90Step 3.7 FlashStepFun25.822026-07-07
91GPT-5.4 nanoOpenAI25.732026-06-01
92DeepSeek-V3DeepSeek25.442026-07-07
93GLM-5.1Z.ai25.222026-06-01
94GLM-4.5Z.ai25.112026-06-01
95Grok 4.1 FastSpaceXAI25.122026-06-23
96Doubao Seed CodeByteDance25.122026-07-07
97Step 3.5 FlashStepFun2542026-07-07
98GPT-5 miniOpenAI2552026-07-07
99Mistral Large 3Mistral24.922026-07-07
100Llama 4 MaverickMeta24.922026-07-07

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