AA-Omniscience

Knowledge recall that penalises hallucination: correct answers minus confident wrong ones, from −100 to 100. Run by Artificial Analysis.

Published by

Artificial Analysis

Category

Knowledge

Index weight

1.0

Models

472

Data as of

9 Sept 2026

Source: Artificial Analysis (artificialanalysis.ai). Data taken from the public leaderboard.

1. 1GPT-6 Astra43.7
2. 2Claude Fable 5.143.5
3. 3GPT-6 Astra43.4
4. 4GPT-6 Astra43.4
5. 5Claude Fable 543.3
6. 6Claude Fable 5.142.4
7. 7GPT-6 Astra42.2
8. 8Claude Fable 5.140.8
9. 9GPT-6 Astra40.5
10. 10Claude Fable 5.137.6
11. 11Claude Opus 537.1
12. 12Claude Opus 535.4
13. 13Claude Fable 5.134.1
14. 14Claude Opus 533.7
15. 15Gemini 3.1 Pro31.9

SearchAll providersAI21 LabsAI9StarsAlibabaAllen Institute for AIAllenaiAmazonAnt GroupAnt-groupAnthropicApodexArcee AIBaiduByteDanceCelerisChina MobileCohereDeep CogitoDeepSeekGoogleIBMInceptionInclusionAIKorea TelecomKwaiKATLG AI ResearchLiquid AILongCatMBZUAI Institute of Foundation ModelsMetaMicrosoftMiniMaxMistral AIMoonshot AIMotif TechnologiesMultiverse ComputingNanbeigeNaverNex AGINous ResearchNVIDIAOpenAIOpenBMBPrime IntellectQwenReka AISapiens AISarvamServiceNowSK TelecomStepFunSwiss AI InitiativeThinking MachinesTII UAETrillion LabsUnknownUpstagexAIXiaomiZhipu AIOpen weights472 of 472

#	Model	Score ↓	Run ↕	Index ↕	Published ↕
1	GPT-6 AstraOpenAI	43.7	–	69.5	–
2	Claude Fable 5.1Anthropic	43.5	Claude Fable 5.1 (max with fallback)	70.0	–
3	GPT-6 AstraOpenAI	43.4	–	68.6	–
4	GPT-6 AstraOpenAI	43.4	GPT-6 Astra (max)	69.9	–
5	Claude Fable 5Anthropic	43.3	Claude Fable 5 (with fallback)	70.4	–
6	Claude Fable 5.1Anthropic	42.4	–	70.3	–
7	GPT-6 AstraOpenAI	42.2	–	67.4	–
8	Claude Fable 5.1Anthropic	40.8	–	69.7	–
9	GPT-6 AstraOpenAI	40.5	–	65.8	–
10	Claude Fable 5.1Anthropic	37.6	–	68.0	–
11	Claude Opus 5Anthropic	37.1	Claude Opus 5 (max)	70.0	–
12	Claude Opus 5Anthropic	35.4	–	66.3	–
13	Claude Fable 5.1Anthropic	34.1	–	65.9	–
14	Claude Opus 5Anthropic	33.7	–	69.3	–
15	Gemini 3.1 ProGoogle	31.9	–	63.9	–
16	Claude Opus 5Anthropic	31.0	–	63.1	–
17	Grok 4.6xAI	30.5	Grok 4.6 (high)	63.6	–
18	Gemini 3.8 FlashGoogle	29.6	Gemini 3.8 Flash (high)	62.4	–
19	Grok 4.6xAI	29.3	–	62.3	–
20	Claude Opus 4.8Anthropic	28.8	Claude Opus 4.8 (max)	64.7	–
21	Gemini 3.8 FlashGoogle	28.6	–	63.5	–
22	Claude Opus 5Anthropic	28.6	–	61.7	–
23	Muse Spark 1.1Meta	28.1	Muse Spark 1.1 (xhigh)	64.9	–
24	Grok 4.6xAI	28	–	63.8	–
25	Claude Opus 4.7Anthropic	27.3	Claude Opus 4.7 (max)	65.5	–
26	Muse Spark 1.2Meta	27.2	Muse Spark 1.2 (xhigh)	62.6	–
27	GPT-6 AstraOpenAI	26.6	–	64.4	–
28	Gemini 3.7 FlashGoogle	26.5	Gemini 3.7 Flash (high)	61.7	–
29	Grok 4.6xAI	25.9	–	59.6	–
30	Grok 4.5xAI	25.3	Grok 4.5 (high)	63.4	–
31	Muse Spark 1.3Meta	25	Muse Spark 1.3 (max)	–	–
32	Gemini 3.7 FlashGoogle	23.7	–	64.4	–
33	Muse Spark 1.3Meta	23.1	–	63.9	–
34	Gemini 3.7 FlashGoogle	22.1	–	62.5	–
35	Gemini 3.6 FlashGoogle	22.1	Gemini 3.6 Flash	60.7	–
36	GPT-5.6 SolOpenAI	22.0	GPT-5.6 Sol (max)	65.3	–
37	Gemini 3.8 FlashGoogle	21.4	–	61.9	–
38	Gemini 3.5 FlashGoogle	21.2	Gemini 3.5 Flash	60.7	–
39	GPT-5.6 SolOpenAI	21.0	–	67.5	–
40	Gemini 3.5 FlashGoogle	20.8	–	63.9	–
41	GPT-5.5OpenAI	20.5	GPT-5.5 (xhigh)	66.6	–
42	GPT-5.6 SolOpenAI	20.4	–	68.0	–
43	Kimi K3Moonshot AIopen	19.7	Kimi K3 (max)	66.5	–
44	GPT-5.6 SolOpenAI	19.4	–	66.0	–
45	GPT-5.6 SolOpenAI	18.9	–	64.6	–
46	GPT-5.5OpenAI	18.8	–	66.3	–
47	GPT-5.5OpenAI	18.1	–	64.9	–
48	Grok 4.3xAI	18.0	Grok 4.3 (high)	57.8	–
49	Grok 4.3xAI	16.7	–	60.7	–
50	Claude Sonnet 5Anthropic	16.4	Claude Sonnet 5 (max)	61.2	–
51	Gemini 3 ProGoogle	15.3	Gemini 3 Pro Preview (high)	61.6	–
52	GPT-5.5OpenAI	15.1	–	62.3	–
53	Claude Opus 4.7Anthropic	14.8	–	60.1	–
54	Grok 4.20xAI	14.8	Grok 4.20 0309 v2	57.9	–
55	GLM-5.3Zhipu AIopen	14.3	GLM-5.3 (max)	–	–
56	Claude Opus 4.5Anthropic	14	–	60.6	–
57	Grok 4.3xAI	13.9	–	59.5	–
58	Claude Opus 4.6Anthropic	13.7	–	63.1	–
59	Qwen3 7Alibaba	13.5	–	58.9	–
60	Claude Sonnet 4.6Anthropic	12.2	–	60.3	–

Show 100 more