Artificial Analysis Intelligence Index v4.3
26 models
Top 5 Models Performance
| OpenAI: gpt-6-astra | ######################################## | 53 |
| Anthropic: claude-fable-5.1 | ######################################## | 53 |
| Anthropic: claude-opus-5 | ###################################### | 51 |
| Anthropic: Claude Fable 5 | ###################################### | 50 |
| Meta: muse-spark-1.3 | #################################### | 48 |
| Rank | Model | Score | |
|---|---|---|---|
| 🇺🇸 | 1 | OpenAI: gpt-6-astra | 53 |
| 🇺🇸 | 2 | Anthropic: claude-fable-5.1 | 53 |
| 🇺🇸 | 3 | Anthropic: claude-opus-5 | 51 |
| 🇺🇸 | 4 | Anthropic: Claude Fable 5 | 50 |
| 🇺🇸 | 5 | Meta: muse-spark-1.3 | 48 |
| 🇺🇸 | 6 | OpenAI: gpt-5.6-sol | 47 |
| 🇨🇳 | 7 | Z.ai: glm-5.3北京智谱华章科技有限公司 | 45 |
| 🇺🇸 | 8 | SpaceXAI: grok-4.6 | 44 |
| 🇨🇳 | 9 | Moonshot AI: kimi-k3月之暗面 | 44 |
| 🇨🇳 | 10 | Z.ai: glm-5.3-flash北京智谱华章科技有限公司 | 42 |
| 🇺🇸 | 11 | OpenAI: gpt-5.6-terra | 42 |
| 🇺🇸 | 12 | Google: gemini-3.8-flash | 41 |
| 🇨🇳 | 13 | Qwen: qwen3.8-2.4t-a95b通义千问 | 40 |
| 🇺🇸 | 14 | OpenAI: gpt-5.6-luna | 38 |
| 🇨🇳 | 15 | DeepSeek: deepseek-v4-pro-0813深度求索 | 36 |
| 🇨🇳 | 16 | DeepSeek: deepseek-v4-flash-0731深度求索 | 35 |
| 🇨🇳 | 17 | Qwen: qwen3.8-27b通义千问 | 34 |
| 🇨🇳 | 18 | MiniMax: MiniMax M3稀宇科技 | 30 |
| 🇺🇸 | 19 | Thinking Machines Lab: inkling | 26 |
| 🇺🇸 | 20 | Google: gemini-3.5-flash-lite | 23 |
| 🇺🇸 | 21 | NVIDIA Corporation: nvidia-nemotron-3-ultra | 23 |
| 🇺🇸 | 22 | Meta: muse-glimmer-30b | 18 |
| 🇫🇷 | 23 | Mistral AI: mistral-medium-3.5-128b | 15 |
| 🇺🇸 | 24 | OpenAI: gpt-oss-120b | 12 |
| 🇺🇸 | 25 | OpenAI: gpt-oss-20b | 9 |
| 🇺🇸 | 26 | OpenAI: GPT-3.5 Turbo | 5 |