Skip to main content
BenchLM
Data

Artificial Analysis Omniscience Accuracy (AA-Omniscience Accuracy)

Data verified

An Artificial Analysis knowledge metric for the proportion of correctly answered questions.

Top models on AA-Omniscience Accuracy — October 29, 2026

As of October 29, 2026, Claude Fable 5.1 leads the AA-Omniscience Accuracy leaderboard with 68.4% , followed by Claude Opus 5.5 (65.5%) and Claude Fable 5 (65.4%).

1Closed

Claude Fable 5.1

Anthropic

68.4%
Overall 81.08Context 1M
2Closed

Claude Opus 5.5

Anthropic

65.5%
Overall 86.29Context 1M
3Closed

Claude Fable 5

Anthropic

65.4%
Overall 80.78Context 1M+

158 modelsKnowledge17% of Knowledge reference weightCurrentUpdated October 29, 2026

Leaderboard (158 models)

AA-Omniscience Accuracy leaderboard
RankModel / configurationScoreParameters (B)Open / closed
1Claude Fable 5.1Anthropic
68.4%
Not reportedClosed
2Claude Opus 5.5Anthropic
65.5%
Not reportedClosed
3Claude Fable 5Anthropic
65.4%
Not reportedClosed
4GPT-6.1 SolOpenAI
61.8%
Not reportedClosed
5GPT-6 AstraOpenAI
61.6%
Not reportedClosed
6Claude Opus 5Anthropic
60.9%
Not reportedClosed
7GPT-5.6 SolOpenAI
59.4%
Not reportedClosed
8Gemini 4 CarbonGoogle
58.3%
Not reportedClosed
9GPT-5.5OpenAI
58.0%
Not reportedClosed
10Gemini 3 ProGoogle
55.8%
Not reportedClosed
11Gemini 3.7 FlashGoogle
55.3%
Not reportedClosed
12GPT-6 SolOpenAI
54.5%
Not reportedClosed
13Muse Spark 1.1Meta
52.1%
Not reportedClosed
14Gemini 3.5 FlashGoogle
51.9%
Not reportedClosed
15Grok 4.5xAI
51.6%
Not reportedClosed
16Claude Sonnet 5.5Anthropic
51.5%
Not reportedClosed
17GPT-5.4OpenAI
50.8%
Not reportedClosed
18Gemini 3.8 FlashGoogle
50.5%
Not reportedClosed
19Gemini 3.6 FlashGoogle
50.0%
Not reportedClosed
20Muse SparkMeta
49.6%
Not reportedClosed
21DeepSeek V4 Pro 0813DeepSeek
49.1%
Not reportedOpen
22Claude Opus 4.7 (Adaptive)Anthropic
48.9%
Not reportedClosed
23Kimi K3.1 ProMoonshot AI
48.9%
Not reportedPending
24Claude Opus 4.8Anthropic
48.8%
Not reportedClosed
25Gemini 4 ArgonGoogle
48.3%
Not reportedClosed
26Grok 4.6xAI
48.2%
Not reportedClosed
27Kimi K3Moonshot AI
47.6%
Not reportedOpen
28GPT-6.1 TerraOpenAI
47.6%
Not reportedClosed
29Claude Opus 4.6 (Adaptive)Anthropic
47.0%
Not reportedClosed
30GPT-5.6 TerraOpenAI
46.8%
Not reportedClosed
31Claude Opus 4.5 ThinkingAnthropic
46.6%
Not reportedClosed
32DeepSeek V4.1 Flash 1017DeepSeek
46.4%
Not reportedOpen
33Claude Opus 4.6Anthropic
45.8%
Not reportedClosed
34Gemini 3 FlashGoogle
45.8%
Not reportedClosed
35Muse Spark 1.2Meta
45.4%
Not reportedClosed
36MiniMax M3.1 FlashMiniMax
45.1%
Not reportedClosed
37MiniMax M3MiniMax
44.9%
Not reportedOpen
38Claude Opus 4.7Anthropic
44.7%
Not reportedClosed
39GPT-5.2OpenAI
44.3%
Not reportedClosed
40Muse Spark 1.3Meta
43.6%
Not reportedClosed
41GPT-5.6 LunaOpenAI
42.7%
Not reportedClosed
42Grok 4.7xAI
42.7%
Not reportedClosed
43Muse Spark OpenMeta
42.0%
Not reportedOpen
44DeepSeek V4 Pro (High)DeepSeek
41.4%
Not reportedOpen
45GPT-5.2-CodexOpenAI
41.1%
Not reportedClosed
46Muse Wonder FullMeta
41.0%
Not reportedClosed
47Claude Opus 4.5Anthropic
40.9%
Not reportedClosed
48Grok 4xAI
40.5%
Not reportedClosed
49DeepSeek V4 Flash 0731DeepSeek
40.4%
Not reportedOpen
50GPT-5 (high)OpenAI
40.3%
Not reportedClosed
51Claude Sonnet 5Anthropic
40.1%
Not reportedClosed
52GPT-5.1-CodexOpenAI
39.9%
Not reportedClosed
53GPT-5.1-Codex-MaxOpenAI
39.9%
Not reportedClosed
54GPT-6 LunaOpenAI
39.6%
Not reportedClosed
55GPT-5 (medium)OpenAI
39.5%
Not reportedClosed
56InklingThinking Machines Lab
39.3%
Not reportedOpen
57Gemini 2.5 ProGoogle
39.1%
Not reportedClosed
58Claude Sonnet 4.6Anthropic
38.6%
Not reportedClosed
59Muse WonderMeta
38.0%
Not reportedClosed
60Ling 3.0 TinyInclusionAI
37.9%
Not reportedOpen
61Qwen 3.6 Max (preview)Alibaba
37.9%
Not reportedClosed
62GPT-5.1OpenAI
37.7%
Not reportedClosed
63GPT-5.4 miniOpenAI
37.5%
Not reportedClosed
64Step 5 PreviewStepFun
36.4%
Not reportedOpen
65Kimi K2.5 (Reasoning)Moonshot AI
35.2%
Not reportedClosed
66Kimi K2.5Moonshot AI
35.2%
Not reportedOpen
67MiMo-V2.6-ProXiaomi
34.8%
Not reportedOpen
68Claude Haiku 5.5Anthropic
34.7%
Not reportedClosed
69Grok 4.3xAI
34.6%
Not reportedClosed
70o1OpenAI
34.5%
Not reportedClosed
71Kimi K2.6Moonshot AI
32.6%
Not reportedOpen
72GLM-5.3Z.AI
32.3%
Not reportedOpen
73Inkling-SmallThinking Machines Lab
31.8%
Not reportedOpen
74Hy3 PreviewTencent
31.5%
Not reportedOpen
75Qwen3.7 MaxAlibaba
31.1%
Not reportedClosed
76Beam 1.0 501B-A23BReflection AI
30.7%
Not reportedOpen
77DeepSeek-R1DeepSeek
30.5%
Not reportedOpen
78Qwen3.8 Max PreviewAlibaba
29.5%
Not reportedClosed
79GLM-4.7Z.AI
29.3%
Not reportedOpen
80GLM-5V-TurboZ.AI
29.3%
Not reportedClosed
81DeepSeek V3.1 (Reasoning)DeepSeek
29.0%
Not reportedOpen
82GLM-5-TurboZ.AI
28.4%
Not reportedClosed
83Claude 3.7 SonnetAnthropic
28.2%
Not reportedClosed
84Muse Glimmer 30BMeta
28.0%
Not reportedOpen
85Claude 3.7 Sonnet ThinkingAnthropic
28.0%
Not reportedClosed
86GPT-4.1OpenAI
27.8%
Not reportedClosed
87Kimi K2Moonshot AI
27.4%
Not reportedClosed
88MiMo-V2.6-FlashXiaomi
27.2%
Not reportedOpen
89MiniMax M2.7MiniMax
26.8%
Not reportedOpen
90MiMo-V2-ProXiaomi
26.6%
Not reportedClosed
91Gemini 3.5 Flash-LiteGoogle
26.5%
Not reportedClosed
92Qwen3.6 PlusAlibaba
26.4%
Not reportedClosed
93GLM-5Z.AI
26.3%
Not reportedOpen
94Nemotron 3.5 Super VL 120B A12BNVIDIA
26.2%
Not reportedOpen
95MiniMax M2.5MiniMax
26.2%
Not reportedClosed
96Gemini 2.5 FlashGoogle
26.1%
Not reportedClosed
97Nemotron 3.5 Super VL 120B A12B NVFP4NVIDIA
25.9%
Not reportedOpen
98Step 3.7 FlashStepFun
25.8%
Not reportedOpen
99GPT-5.4 nanoOpenAI
25.7%
Not reportedClosed
100DeepSeek V3DeepSeek
25.5%
Not reportedOpen
101Grok 4.1 Fast (Reasoning)xAI
25.1%
Not reportedClosed
102Mistral Large 3Mistral
25.0%
Not reportedClosed
103GPT-5 miniOpenAI
25.0%
Not reportedClosed
104o4-miniOpenAI
24.8%
Not reportedClosed
105Step 3.5 FlashStepFun
24.6%
Not reportedOpen
106Qwen3 MaxAlibaba
24.4%
Not reportedClosed
107GLM-5.2Z.AI
24.3%
Not reportedOpen
108DeepSeek V3 0324DeepSeek
24.3%
Not reportedOpen
109DeepSeek V3.2DeepSeek
24.0%
Not reportedOpen
110Mistral Large 4Mistral
23.9%
Not reportedOpen
111GLM-5.1Z.AI
23.7%
Not reportedOpen
112Qwen3.8-Flash-NextAlibaba
23.5%
Not reportedOpen
113Grok Code Fast 1xAI
23.5%
Not reportedClosed
114Gemini 2.0 FlashGoogle
23.5%
Not reportedClosed
115DeepSeek V3.1DeepSeek
23.1%
Not reportedOpen
116Mistral Medium 3.5 128BMistral
23.0%
Not reportedOpen
117Grok 4 Fast (Reasoning)xAI
22.8%
Not reportedClosed
118Claude 4 SonnetAnthropic
22.7%
Not reportedClosed
119MiniMax M1 80kMiniMax
22.5%
Not reportedClosed
120MiMo-V2.5-ProXiaomi
22.4%
Not reportedClosed
121Mercury 2.5Inception
22.0%
Not reportedClosed
122Nemotron 3.5 Bolt 74B-A9B BF16NVIDIA
21.8%
Not reportedOpen
123Nemotron 3 UltraNVIDIA
21.6%
Not reportedOpen
124Mistral Small 4Mistral
21.4%
Not reportedOpen
125GLM-4.6Z.AI
21.4%
Not reportedOpen
126Mercury 2Inception
21.2%
Not reportedClosed
127Qwen3.5-27BAlibaba
20.7%
Not reportedOpen
128GPT-4.1 miniOpenAI
20.3%
Not reportedClosed
129Magistral MediumMistral
20.3%
Not reportedClosed
130Qwen3.5-35B-A3BAlibaba
20.1%
Not reportedOpen
131GPT-4oOpenAI
19.9%
Not reportedClosed
132Mistral Large 2Mistral
19.9%
Not reportedClosed
133Qwen3.6-27BAlibaba
19.6%
Not reportedOpen
134MiMo-V2-OmniXiaomi
19.3%
Not reportedClosed
135GPT-5 nanoOpenAI
19.1%
Not reportedClosed
136Solar Pro 3Upstage
18.5%
Not reportedClosed
137Mistral Medium 3Mistral
18.3%
Not reportedClosed
138Ling 3.0 FlashInclusionAI
18.2%
Not reportedOpen
139Magistral SmallMistral
18.0%
Not reportedOpen
140Claude 3 HaikuAnthropic
17.6%
Not reportedClosed
141Qwen3.8-27BAlibaba
17.3%
Not reportedOpen
142Grok 4.1 FastxAI
17.2%
Not reportedClosed
143Nova ProAmazon
16.9%
Not reportedClosed
144K-ExaoneLG AI Research
16.4%
Not reportedClosed
145GLM-4.5-AirZ.AI
16.3%
Not reportedClosed
146GLM-4.7-FlashZ.AI
16.2%
Not reportedOpen
147Solar Pro 2Upstage
16.1%
Not reportedClosed
148MiMo-V2-FlashXiaomi
15.6%
Not reportedOpen
149Ling 2.6 FlashInclusionAI
15.6%
Not reportedOpen
150Gemini 2.5 Flash-LiteGoogle
15.4%
Not reportedClosed
151Mistral Small 3.1Mistral
15.1%
Not reportedOpen
152Devstral SmallMistral
14.7%
Not reportedOpen
153GPT-4.1 nanoOpenAI
13.7%
Not reportedClosed
154Claude 3.5 HaikuAnthropic
13.2%
Not reportedClosed
155Gemma 3 27BGoogle
13.0%
Not reportedOpen
156Command A+Cohere
9.8%
Not reportedOpen
157LFM2.5-8B-A1BLiquidAI
9.4%
Not reportedOpen
158LFM3-1.7BLiquidAI
7.1%
Not reportedOpen

According to BenchLM.ai, Claude Fable 5.1 leads the AA-Omniscience Accuracy benchmark with a score of 68.4%, followed by Claude Opus 5.5 (65.5%) and Claude Fable 5 (65.4%). The top models are clustered within 3.0 points, suggesting this benchmark is nearing saturation for frontier models.

158 models have been evaluated on AA-Omniscience Accuracy. The benchmark falls in the Knowledge category. BenchAlign v5.8 gives AA-Omniscience Accuracy 17% of the Knowledge reference weight, so it moves the Knowledge leaderboard and the overall ranking. Reference weights are relative weights in the calibrated model, not fixed shares of a score.

About AA-Omniscience Accuracy

Year

2026

Tasks

Knowledge questions

Format

Accuracy

Difficulty

Broad knowledge

BenchLM stores AA-Omniscience Accuracy when a model page publishes the exact Artificial Analysis benchmark card value.

Artificial Analysis model benchmarks

Freshness and provenance

Version

AA-Omniscience Accuracy 2026

Refresh cadence

Quarterly

Staleness state

Current

Question availability

Public benchmark set

Current

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 AA-Omniscience Accuracy measure?

An Artificial Analysis knowledge metric for the proportion of correctly answered questions.

Which model scores highest on AA-Omniscience Accuracy?

Claude Fable 5.1 by Anthropic currently leads with a score of 68.4% on AA-Omniscience Accuracy.

How many models are evaluated on AA-Omniscience Accuracy?

158 AI models have published results on AA-Omniscience Accuracy in the BenchLM catalog.

Compare top models on AA-Omniscience Accuracy

Claude Fable 5.1 vs Claude Opus 5.5Claude Opus 5.5 vs Claude Fable 5Claude Fable 5 vs GPT-6.1 SolGPT-6.1 Sol vs GPT-6 Astra
Last updated: October 29, 2026 · BenchLM version AA-Omniscience Accuracy 2026

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 6,100+ subscribers.

One email each week. Unsubscribe anytime.