BENCHMARKS, WITH CONTEXT
モデルの実力は、データで確かめる。
看见评分，也看见来源、方法和运行配置。按已采集的公开评测数据排序。
来源：Artificial Analysis 公开模型页 ↗。当前 31 条匹配评测；仅代表已采集集合。不同推理预算、提示配置与版本不应混为同一模型评分。
| 順位 | モデル・提供元 | AA知能指数 | 実行設定 | 出典・収集日時 |
|---|---|---|---|---|
| 1 | Gemini 4 Carbon (High)Google | 59.43 | High | Artificial Analysis ↗2026/11/2 06:30:46 |
| 2 | Claude Opus 5.5 (Max, Default Fallback)Anthropic | 58.41 | Max, Default Fallback | Artificial Analysis ↗2026/11/2 06:30:46 |
| 3 | GPT-6 Astra (Max)OpenAI | 56.85 | Max | Artificial Analysis ↗2026/11/2 06:30:46 |
| 4 | GPT-6.1 Sol (Max)OpenAI | 54.51 | Max | Artificial Analysis ↗2026/11/2 06:30:46 |
| 5 | Claude Sonnet 5.5 (Max, Default Fallback)Anthropic | 54.24 | Max, Default Fallback | Artificial Analysis ↗2026/11/2 06:30:46 |
| 6 | GPT-6.1 Terra (Max)OpenAI | 50.98 | Max | Artificial Analysis ↗2026/11/2 06:30:46 |
| 7 | Gemini 4 ArgonGoogle | 50.41 | High | Artificial Analysis ↗2026/10/28 23:39:06 |
| 8 | Muse Wonder Full (Max)Meta | 49.73 | Max | Artificial Analysis ↗2026/11/2 06:30:46 |
| 9 | Claude Fable 5.1 (Max, Default Fallback)Anthropic | 49.01 | Max, Default Fallback | Artificial Analysis ↗2026/11/2 06:30:46 |
| 10 | Kimi K3.1 ProKimi | 46.00 | Kimi K3.1 Pro | Artificial Analysis ↗2026/11/2 06:30:46 |
| 11 | Muse Wonder (Max)Meta | 43.32 | Max | Artificial Analysis ↗2026/11/2 06:30:46 |
| 12 | Claude Haiku 5.5 (Max)Anthropic | 36.23 | Max | Artificial Analysis ↗2026/11/2 06:30:46 |
| 13 | Grok 4.7 (Xhigh)SpaceXAI | 34.34 | Xhigh | Artificial Analysis ↗2026/11/2 06:30:46 |
| 14 | Qwen3.8 Max (0902)阿里巴巴 · Qwen | 33.55 | Qwen3.8 Max (0902) | Artificial Analysis ↗2026/11/2 06:30:46 |
| 15 | DeepSeek V4.1 Flash 1017 (Max)DeepSeek | 32.49 | Max | Artificial Analysis ↗2026/11/2 06:30:46 |
| 16 | GLM-5.3 (Max)Z AI | 31.65 | Max | Artificial Analysis ↗2026/11/2 06:30:46 |
| 17 | Gemini 3.8 Flash (High)Google | 30.55 | High | Artificial Analysis ↗2026/11/2 06:30:46 |
| 18 | MiMo-V2.6-ProXiaomi | 29.89 | MiMo-V2.6-Pro | Artificial Analysis ↗2026/11/2 06:30:46 |
| 19 | Ling 3.1 Flash 1019InclusionAI | 29.47 | Ling 3.1 Flash 1019 | Artificial Analysis ↗2026/11/2 06:30:46 |
| 20 | MiniMax-M3.1-FlashMiniMax | 28.91 | MiniMax-M3.1-Flash | Artificial Analysis ↗2026/11/2 06:30:46 |
| 21 | Step 5 Preview阶跃星辰 | 28.53 | Step 5 Preview | Artificial Analysis ↗2026/11/2 06:30:46 |
| 22 | GPT-6 Luna (Max)OpenAI | 27.76 | Max | Artificial Analysis ↗2026/11/2 06:30:46 |
| 23 | Mistral Large 4 (High)Mistral AI | 26.02 | High | Artificial Analysis ↗2026/11/2 06:30:46 |
| 24 | Mistral Large 4 PreviewMistral AI | 25.29 | Mistral Large 4 Preview | Artificial Analysis ↗2026/10/29 17:45:04 |
| 25 | Mistral Large 4 (High)Mistral AI | 25.29 | Mistral Large 4 Preview | Artificial Analysis ↗2026/10/29 11:41:31 |
| 26 | Muse Spark Open (Xhigh)Meta | 24.68 | Xhigh | Artificial Analysis ↗2026/11/2 06:30:46 |
| 27 | Qwen3.8 27B (Xhigh)阿里巴巴 · Qwen | 20.11 | Xhigh | Artificial Analysis ↗2026/11/2 06:30:46 |
| 28 | K2 Horizon 375B A23BInstitute of Foundation Models | 17.77 | K2 Horizon 375B A23B | Artificial Analysis ↗2026/11/2 06:30:46 |
| 29 | Beam 1.0Reflection | 17.73 | Beam 1.0 | Artificial Analysis ↗2026/11/2 06:30:46 |
| 30 | Nemotron 3.5 Super VLNVIDIA | 13.62 | Nemotron 3.5 Super VL | Artificial Analysis ↗2026/11/2 06:30:46 |
| 31 | Inkling (Xhigh)Thinking Machines | 12.48 | Xhigh | Artificial Analysis ↗2026/11/2 06:30:46 |
ランキングの見方
分数来自第三方公开测试，不是本站实测。各测试覆盖的能力、提示和推理预算不同。建议点击模型查看来源，并使用自己的业务数据进行验证。采集失败时保留上一份快照及其时间，不编造新评分。