Artificial Analysis Intelligence Index
We show this table for reference; we do not rank on it.
Gemini 4 Carbon has the highest reported Artificial Analysis Intelligence Index score at 59.4%, ahead of GPT-5.6 Sol (58.9%) and Claude Opus 5.5 (58.4%) among 170 models. We compile the table from provider self-reports and secondary reports and keep it for display only; it does not affect overall rankings.
Artificial Analysis describes itself as an independent AI benchmarking company. Its Intelligence Index combines the company's own evaluations of AI models into one score per model. BenchLM compiles reported index scores, which can come from different index versions, and shows them for reference only.
Benchmark score on Artificial Analysis Intelligence Index — October 29, 2026
We compile the Artificial Analysis Intelligence Index rows from provider self-reports and secondary reports. Gemini 4 Carbon leads the table at 59.4%, followed by GPT-5.6 Sol (58.9%) and Claude Opus 5.5 (58.4%). We do not use these results to rank models overall.
Gemini 4 Carbon
GPT-5.6 Sol
OpenAI
Claude Opus 5.5
Anthropic
170 modelsKnowledgeCurrentDisplay onlyUpdated October 29, 2026
Benchmark score table (170 models)
| Rank | Model / configuration | Score | Parameters (B) | Open / closed |
|---|---|---|---|---|
| 1 | Gemini 4 CarbonGoogle | 59.4% | Not reported | Closed |
| 2 | GPT-5.6 SolOpenAI | 58.9% | Not reported | Closed |
| 3 | Claude Opus 5.5Anthropic | 58.4% | Not reported | Closed |
| 4 | GPT-6 AstraOpenAI | 56.9% | Not reported | Closed |
| 5 | GPT-5.6 TerraOpenAI | 55.0% | Not reported | Closed |
| 6 | GPT-6.1 SolOpenAI | 54.5% | Not reported | Closed |
| 7 | Claude Sonnet 5.5Anthropic | 54.2% | Not reported | Closed |
| 8 | DeepSeek V4 Pro 0813DeepSeek | 53.2% | Not reported | Open |
| 9 | GPT-5.6 LunaOpenAI | 51.2% | Not reported | Closed |
| 10 | GPT-6.1 TerraOpenAI | 51.0% | Not reported | Closed |
| 11 | Claude Opus 5Anthropic | 50.8% | Not reported | Closed |
| 12 | Gemini 4 ArgonGoogle | 50.4% | Not reported | Closed |
| 13 | Gemini 3.5 FlashGoogle | 50.2% | Not reported | Closed |
| 14 | Muse Wonder FullMeta | 49.7% | Not reported | Closed |
| 15 | Claude Fable 5Anthropic | 49.6% | Not reported | Closed |
| 16 | Claude Fable 5.1Anthropic | 49.0% | Not reported | Closed |
| 17 | Muse Spark 1.3Meta | 48.1% | Not reported | Closed |
| 18 | GPT-6 SolOpenAI | 47.6% | Not reported | Closed |
| 19 | MiMo-V2.6-ProXiaomi | 46.3% | Not reported | Open |
| 20 | Kimi K3.1 ProMoonshot AI | 46.0% | Not reported | Pending |
| 21 | Grok 4.6xAI | 44.3% | Not reported | Closed |
| 22 | DeepSeek V4 Pro (High)DeepSeek | 43.8% | Not reported | Open |
| 23 | Kimi K3Moonshot AI | 43.6% | Not reported | Open |
| 24 | Muse WonderMeta | 43.3% | Not reported | Closed |
| 25 | Claude Opus 4.8Anthropic | 41.8% | Not reported | Closed |
| 26 | Hy3 PreviewTencent | 41.2% | Not reported | Open |
| 27 | Claude Opus 4.7 (Adaptive)Anthropic | 40.7% | Not reported | Closed |
| 28 | Muse Spark 1.2Meta | 39.6% | Not reported | Closed |
| 29 | DeepSeek V4.1 Flash 1017DeepSeek | 39.5% | Not reported | Open |
| 30 | Gemini 3.7 FlashGoogle | 39.1% | Not reported | Closed |
| 31 | GPT-5.4OpenAI | 39.0% | Not reported | Closed |
| 32 | Grok 4.5xAI | 38.8% | Not reported | Closed |
| 33 | GPT-5.5OpenAI | 38.4% | Not reported | Closed |
| 34 | Claude Sonnet 5Anthropic | 38.2% | Not reported | Closed |
| 35 | Grok 4.3xAI | 37.6% | Not reported | Closed |
| 36 | Claude Haiku 5.5Anthropic | 36.2% | Not reported | Closed |
| 37 | Grok 4.7xAI | 34.3% | Not reported | Closed |
| 38 | DeepSeek V4 Flash 0731DeepSeek | 34.3% | Not reported | Open |
| 39 | Gemini 3.6 FlashGoogle | 34.0% | Not reported | Closed |
| 40 | Muse Spark 1.1Meta | 33.7% | Not reported | Closed |
| 41 | GLM-5.2Z.AI | 33.7% | Not reported | Open |
| 42 | Qwen3.8 Max PreviewAlibaba | 33.5% | Not reported | Closed |
| 43 | Claude Opus 4.6 (Adaptive)Anthropic | 31.9% | Not reported | Closed |
| 44 | GLM-5.3Z.AI | 31.6% | Not reported | Open |
| 45 | Muse SparkMeta | 31.3% | Not reported | Closed |
| 46 | Claude Opus 4.7Anthropic | 30.9% | Not reported | Closed |
| 47 | Gemini 3.8 FlashGoogle | 30.6% | Not reported | Closed |
| 48 | GPT-5.2OpenAI | 30.4% | Not reported | Closed |
| 49 | Qwen3.7 MaxAlibaba | 29.5% | Not reported | Closed |
| 50 | Claude Opus 4.5 ThinkingAnthropic | 29.1% | Not reported | Closed |
| 51 | MiniMax M3.1 FlashMiniMax | 28.9% | Not reported | Closed |
| 52 | MiMo-V2-ProXiaomi | 28.6% | Not reported | Closed |
| 53 | Step 5 PreviewStepFun | 28.5% | Not reported | Open |
| 54 | GPT-5.2-CodexOpenAI | 28.5% | Not reported | Closed |
| 55 | Qwen 3.6 Max (preview)Alibaba | 28.4% | Not reported | Closed |
| 56 | Gemini 3 ProGoogle | 28.0% | Not reported | Closed |
| 57 | GLM-5Z.AI | 27.9% | Not reported | Open |
| 58 | GPT-6 LunaOpenAI | 27.8% | Not reported | Closed |
| 59 | Qwen3.6 PlusAlibaba | 27.0% | Not reported | Closed |
| 60 | Kimi K2.6Moonshot AI | 27.0% | Not reported | Open |
| 61 | MiMo-V2.6-FlashXiaomi | 26.7% | Not reported | Open |
| 62 | GLM-5-TurboZ.AI | 26.6% | Not reported | Closed |
| 63 | Claude Opus 4.6Anthropic | 26.4% | Not reported | Closed |
| 64 | GLM-5.1Z.AI | 26.1% | Not reported | Open |
| 65 | MiMo-V2.5-ProXiaomi | 26.0% | Not reported | Closed |
| 66 | Mistral Large 4Mistral | 25.3% | Not reported | Open |
| 67 | GPT-5.1OpenAI | 24.7% | Not reported | Closed |
| 68 | Claude Sonnet 4.6Anthropic | 24.7% | Not reported | Closed |
| 69 | Muse Spark OpenMeta | 24.7% | Not reported | Open |
| 70 | GPT-5.4 miniOpenAI | 24.1% | Not reported | Closed |
| 71 | MiMo-V2-OmniXiaomi | 23.9% | Not reported | Closed |
| 72 | GPT-5.1-CodexOpenAI | 23.7% | Not reported | Closed |
| 73 | GPT-5.1-Codex-MaxOpenAI | 23.7% | Not reported | Closed |
| 74 | Claude Opus 4.5Anthropic | 23.7% | Not reported | Closed |
| 75 | GLM-5V-TurboZ.AI | 23.5% | Not reported | Closed |
| 76 | Kimi K2.5 (Reasoning)Moonshot AI | 23.5% | Not reported | Closed |
| 77 | Kimi K2.5Moonshot AI | 23.5% | Not reported | Open |
| 78 | GPT-5 (high)OpenAI | 23.0% | Not reported | Closed |
| 79 | Qwen3.5-27BAlibaba | 22.9% | Not reported | Open |
| 80 | GPT-5 (medium)OpenAI | 22.9% | Not reported | Closed |
| 81 | Claude 4.1 Opus ThinkingAnthropic | 22.9% | Not reported | Closed |
| 82 | MiniMax M2.5MiniMax | 22.8% | Not reported | Closed |
| 83 | MiniMax M2.7MiniMax | 22.8% | Not reported | Open |
| 84 | Command A+Cohere | 22.5% | Not reported | Open |
| 85 | Grok 4xAI | 22.5% | Not reported | Closed |
| 86 | GLM-4.7Z.AI | 22.2% | Not reported | Open |
| 87 | o3-proOpenAI | 21.9% | Not reported | Closed |
| 88 | Qwen3.6-27BAlibaba | 21.4% | Not reported | Open |
| 89 | GPT-5.4 nanoOpenAI | 20.7% | Not reported | Closed |
| 90 | Claude 4 Opus ThinkingAnthropic | 20.6% | Not reported | Closed |
| 91 | Grok 4.1 Fast (Reasoning)xAI | 20.4% | Not reported | Closed |
| 92 | Ling 3.0 FlashInclusionAI | 20.1% | Not reported | Open |
| 93 | Qwen3.8-27BAlibaba | 20.1% | Not reported | Open |
| 94 | Step 3.7 FlashStepFun | 19.5% | Not reported | Open |
| 95 | Qwen3.5-35B-A3BAlibaba | 19.3% | Not reported | Open |
| 96 | MiniMax M3MiniMax | 18.7% | Not reported | Open |
| 97 | Claude 4.1 OpusAnthropic | 18.6% | Not reported | Closed |
| 98 | Grok 4 Fast (Reasoning)xAI | 17.9% | Not reported | Closed |
| 99 | Gemini 3 FlashGoogle | 17.9% | Not reported | Closed |
| 100 | Beam 1.0 501B-A23BReflection AI | 17.7% | Not reported | Open |
| 101 | Claude 3.7 Sonnet ThinkingAnthropic | 17.7% | Not reported | Closed |
| 102 | Step 3.5 FlashStepFun | 17.0% | Not reported | Open |
| 103 | GPT-5 miniOpenAI | 16.8% | Not reported | Closed |
| 104 | o4-miniOpenAI | 16.6% | Not reported | Closed |
| 105 | Claude 4 OpusAnthropic | 16.6% | Not reported | Closed |
| 106 | Claude 4 SonnetAnthropic | 16.6% | Not reported | Closed |
| 107 | Gemini 2.5 ProGoogle | 16.1% | Not reported | Closed |
| 108 | MiMo-V2-FlashXiaomi | 16.1% | Not reported | Open |
| 109 | DeepSeek V3.2DeepSeek | 16.0% | Not reported | Open |
| 110 | Qwen3 MaxAlibaba | 15.6% | Not reported | Closed |
| 111 | Claude 3.7 SonnetAnthropic | 15.3% | Not reported | Closed |
| 112 | o1OpenAI | 15.2% | Not reported | Closed |
| 113 | GLM-4.6Z.AI | 14.9% | Not reported | Open |
| 114 | GLM-4.7-FlashZ.AI | 14.9% | Not reported | Open |
| 115 | K-ExaoneLG AI Research | 14.4% | Not reported | Closed |
| 116 | Grok Code Fast 1xAI | 14.1% | Not reported | Closed |
| 117 | Ling 2.6 FlashInclusionAI | 14.1% | Not reported | Open |
| 118 | DeepSeek V3.1DeepSeek | 13.7% | Not reported | Open |
| 119 | Nemotron 3.5 Super VL 120B A12BNVIDIA | 13.6% | Not reported | Open |
| 120 | Nemotron 3.5 Super VL 120B A12B NVFP4NVIDIA | 13.6% | Not reported | Open |
| 121 | DeepSeek V3.1 (Reasoning)DeepSeek | 13.5% | Not reported | Open |
| 122 | DeepSeek-R1DeepSeek | 13.1% | Not reported | Open |
| 123 | GPT-5 nanoOpenAI | 13.0% | Not reported | Closed |
| 124 | Inkling-SmallThinking Machines Lab | 12.9% | Not reported | Open |
| 125 | Kimi K2Moonshot AI | 12.7% | Not reported | Closed |
| 126 | GPT-4.1OpenAI | 12.7% | Not reported | Closed |
| 127 | InklingThinking Machines Lab | 12.5% | Not reported | Open |
| 128 | o3-miniOpenAI | 12.5% | Not reported | Closed |
| 129 | o1-proOpenAI | 12.4% | Not reported | Closed |
| 130 | Nemotron 3 UltraNVIDIA | 11.8% | Not reported | Open |
| 131 | MiniMax M1 80kMiniMax | 11.7% | Not reported | Closed |
| 132 | Gemini 3.5 Flash-LiteGoogle | 11.7% | Not reported | Closed |
| 133 | o1-previewOpenAI | 11.4% | Not reported | Closed |
| 134 | Grok 4.1 FastxAI | 11.3% | Not reported | Closed |
| 135 | GLM-4.5-AirZ.AI | 11.1% | Not reported | Closed |
| 136 | Mercury 2Inception | 10.8% | Not reported | Closed |
| 137 | GPT-4.1 miniOpenAI | 10.2% | Not reported | Closed |
| 138 | Gemini 2.5 FlashGoogle | 9.8% | Not reported | Closed |
| 139 | o1-miniOpenAI | 9.8% | Not reported | Closed |
| 140 | DeepSeek V3 0324DeepSeek | 9.7% | Not reported | Open |
| 141 | GPT-4.5OpenAI | 9.6% | Not reported | Closed |
| 142 | QwQ-32BAlibaba | 9.5% | Not reported | Open |
| 143 | Mistral Large 3Mistral | 9.3% | Not reported | Closed |
| 144 | Mistral Medium 3Mistral | 9.1% | Not reported | Closed |
| 145 | Magistral MediumMistral | 9.1% | Not reported | Closed |
| 146 | Gemini 2.0 FlashGoogle | 8.9% | Not reported | Closed |
| 147 | Claude 3.5 HaikuAnthropic | 8.9% | Not reported | Closed |
| 148 | Claude 3 OpusAnthropic | 8.7% | Not reported | Closed |
| 149 | DeepSeek V3DeepSeek | 8.5% | Not reported | Open |
| 150 | GPT-4oOpenAI | 8.4% | Not reported | Closed |
| 151 | DeepSeek R1 Distill Qwen 32BDeepSeek | 8.4% | Not reported | Open |
| 152 | Magistral SmallMistral | 8.2% | Not reported | Open |
| 153 | Gemini 1.5 ProGoogle | 7.9% | Not reported | Closed |
| 154 | Solar Pro 3Upstage | 7.8% | Not reported | Closed |
| 155 | GPT-4.1 nanoOpenAI | 7.8% | Not reported | Closed |
| 156 | Devstral SmallMistral | 7.6% | Not reported | Open |
| 157 | Mistral Large 2Mistral | 7.6% | Not reported | Closed |
| 158 | Mistral Small 3.1Mistral | 7.1% | Not reported | Open |
| 159 | Gemini 1.5 FlashGoogle | 7.1% | Not reported | Closed |
| 160 | GPT-4 TurboOpenAI | 7.0% | Not reported | Closed |
| 161 | Solar Pro 2Upstage | 7.0% | Not reported | Closed |
| 162 | Nova ProAmazon | 7.0% | Not reported | Closed |
| 163 | Qwen2.5 Coder 32B InstructAlibaba | 6.7% | Not reported | Open |
| 164 | Gemini 2.5 Flash-LiteGoogle | 6.7% | Not reported | Closed |
| 165 | GPT-4o miniOpenAI | 6.7% | Not reported | Closed |
| 166 | DeepSeek-V2.5DeepSeek | 6.6% | Not reported | Open |
| 167 | Muse Glimmer 30BMeta | 6.3% | Not reported | Open |
| 168 | Claude 3 HaikuAnthropic | 5.6% | Not reported | Closed |
| 169 | Gemini 1.0 ProGoogle | 5.3% | Not reported | Closed |
| 170 | Gemma 3 27BGoogle | 4.8% | Not reported | Open |
Among the reported Artificial Analysis Intelligence Index rows, Gemini 4 Carbon is first at 59.4%. The third row is 1.0 points behind. The broader top-10 range is 8.5 points, so many of the published results sit in a relatively narrow band.
170 models have been evaluated on Artificial Analysis Intelligence Index. The benchmark falls in the Knowledge category. Artificial Analysis Intelligence Index is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.
About Artificial Analysis Intelligence Index
Year
2026
Tasks
Cross-benchmark intelligence index
Format
Aggregated model score
Difficulty
Display-only external reference
BenchLM tracks Artificial Analysis as a display-only external reference rather than a weighted benchmark. It is useful as a market snapshot, but it is not a benchmark-native row with a single public task set, scoring harness, or exact-source methodology aligned to BenchLM's core benchmark pages.
Freshness and provenance
Version
Artificial Analysis Intelligence Index 2026
Refresh cadence
Quarterly
Staleness state
Current
Question availability
Public benchmark set
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 Artificial Analysis Intelligence Index measure?
Artificial Analysis describes itself as an independent AI benchmarking company. Its Intelligence Index combines the company's own evaluations of AI models into one score per model. BenchLM compiles reported index scores, which can come from different index versions, and shows them for reference only.
Which model scores highest on Artificial Analysis Intelligence Index?
Gemini 4 Carbon by Google currently leads with a score of 59.4% on Artificial Analysis Intelligence Index.
How many models are evaluated on Artificial Analysis Intelligence Index?
170 AI models have published results on Artificial Analysis Intelligence Index in the BenchLM catalog.
Compare top models on Artificial Analysis Intelligence Index
Know when it’s worth switching models
The model to choose, the cheaper alternative, and the release we would wait on.
Read a sample issueJoin 6,100+ subscribers.
One email each week. Unsubscribe anytime.