New: GPT-6.2 benchmark report is here.Read the report ↗

Best LLMs on Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index

Composite evaluation of knowledge work, coding, science, and general knowledge.

Updated: Data as of:
Widely usedSaturation: not applicable
Official benchmark
Artificial Analysis Intelligence Index results. Each row identifies its benchmark version and evaluation track. Scores from different versions are not directly comparable. Best effort from the latest stored report date per model and version.
Model / creatorVersionReported score
Claude Fable 5.2Anthropic · Proprietary5.0.1 · Independent65 points
GPT-6.2 AstraOpenAI · Proprietary5.0.1 / OpenAI GPT-6.2 launch62.1 points
GPT-6.2 AstraOpenAI · Proprietary5.0.1 · Independent62 points
Gemini 4 CarbonGoogle · Proprietary5.0.1 · Independent60 points
Claude Opus 5.5Anthropic · Proprietary5.0.1 · Independent59 points
GPT-6.2 SolOpenAI · Proprietary5.0.1 / OpenAI GPT-6.2 launch57.3 points
GPT-6 AstraOpenAI · Proprietary5.0.1 · Independent57 points
GPT-6.2 SolOpenAI · Proprietary5.0.1 · Independent57 points
GPT-6.1 SolOpenAI · Proprietary5.0.1 · Independent55 points
Kimi K3.1 ProMoonshot AI · Open source / open weights5.0.1 · Independent46 points
Claude Haiku 5.5Anthropic · Proprietary5.0.1 · Independent36 points
DeepSeek V4.1 Flash 1017DeepSeek · Open source / open weights5.0.1 · Independent32 points
GLM-5.3Z.ai · Open source / open weights5.0.1 · Independent32 points
View all 21 published reports
5.0.1 · Independent

fallback: default · reasoning effort: max

5.0.1 · Independent

fallback: default · reasoning effort: max

GPT-6 Astra57 points
5.0.1 · Independent

reasoning effort: max

GPT-6.2 Sol57 points
5.0.1 · Independent

reasoning effort: max

GPT-6.1 Sol55 points
5.0.1 · Independent

reasoning effort: max

Kimi K3.1 Pro46 points
5.0.1 · Independent

reasoning effort: reported setting

GLM-5.332 points
5.0.1 · Independent

reasoning effort: max

GPT-6.2 Sol57.3 points
5.0.1 / OpenAI GPT-6.2 launch

cost per task usd: 1.01 · reasoning effort: max

OpenAI ↗
GPT-6.2 Sol55.7 points
5.0.1 / OpenAI GPT-6.2 launch

cost per task usd: 0.65 · reasoning effort: xhigh

OpenAI ↗
GPT-6.2 Sol54.9 points
5.0.1 / OpenAI GPT-6.2 launch

cost per task usd: 0.56 · reasoning effort: high

OpenAI ↗
GPT-6.2 Sol52.7 points
5.0.1 / OpenAI GPT-6.2 launch

cost per task usd: 0.41 · reasoning effort: medium

OpenAI ↗
GPT-6.2 Sol40.6 points
5.0.1 / OpenAI GPT-6.2 launch

cost per task usd: 0.18 · reasoning effort: low

OpenAI ↗