第 1 名
GPT-5.6 Sol (xhigh)
OpenAI
能力得分
78.3
63 tok/s
输出速度
全部天梯10 / 50 49 个模型
当前数据AIUtil 能力天梯
模型编码能力天梯
比较模型在软件工程、代码生成与问题修复任务中的表现。
本次数据
收录 50 个模型
当前第一
GPT-5.6 Sol (xhigh)
最高分
78.3
入榜数量
50
最近发布
2026/08/05
前三名
领先梯队
分数越高，排名越靠前
第 2 名
Claude Opus 5 (Adaptive Reasoning, Max Effort)
Anthropic
能力得分
78.0
54 tok/s
输出速度
第 3 名
GPT-5.6 Sol (max)
OpenAI
能力得分
77.4
64 tok/s
输出速度
分数对比
前十名能力差距
- 01GPT-5.6 Sol (xhigh)78.3
- 02Claude Opus 5 (Adaptive Reasoning, Max Effort)78.0
- 03GPT-5.6 Sol (max)77.4
- 04GPT-5.6 Sol (high)77.2
- 05Claude Opus 5 (Adaptive Reasoning, Xhigh Effort)77.0
- 06GPT-5.6 Terra (max)76.7
- 07Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)76.5
- 08Claude Opus 5 (Adaptive Reasoning, High Effort)76.5
- 09GPT-5.6 Sol (medium)76.3
- 10Kimi K3 (max)76.2
020406080Artificial Analysis Coding Index
价格对比
能力与输出价格
越靠近左上角，代表更高能力与更低输出价格。
每百万输出 Token 价格（美元，对数轴）Artificial Analysis Coding Index
80.774.668.562.456.3$0.28$1.02$3.74$14$50
GPT-5.6 Sol (xhigh)
OpenAI
能力分78.3输出价格$30 / 1M
63 tok/s全部结果
完整排名
Artificial Analysis Coding Index
| 排名 | 模型 | 厂商 | Artificial Analysis Coding Index | 输出速度 | 发布日期 |
|---|---|---|---|---|---|
| #01 | GPT-5.6 Sol (xhigh) | OpenAI | 78.3 | 63 tok/s | |
| #02 | Claude Opus 5 (Adaptive Reasoning, Max Effort) | Anthropic | 78.0 | 54 tok/s | |
| #03 | GPT-5.6 Sol (max) | OpenAI | 77.4 | 64 tok/s | |
| #04 | GPT-5.6 Sol (high) | OpenAI | 77.2 | 61 tok/s | |
| #05 | Claude Opus 5 (Adaptive Reasoning, Xhigh Effort) | Anthropic | 77.0 | 55 tok/s | |
| #06 | GPT-5.6 Terra (max) | OpenAI | 76.7 | 125 tok/s | |
| #07 | Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) | Anthropic | 76.5 | 72 tok/s | |
| #08 | Claude Opus 5 (Adaptive Reasoning, High Effort) | Anthropic | 76.5 | 53 tok/s | |
| #09 | GPT-5.6 Sol (medium) | OpenAI | 76.3 | 58 tok/s | |
| #10 | Kimi K3 (max) | Kimi | 76.2 | 38 tok/s | |
| #11 | GPT-5.5 (xhigh) | OpenAI | 74.9 | 76 tok/s | |
| #12 | Claude Opus 4.8 (Adaptive Reasoning, Max Effort) | Anthropic | 74.3 | 58 tok/s | |
| #13 | Claude Opus 5 (Adaptive Reasoning, Medium Effort) | Anthropic | 74.3 | 54 tok/s | |
| #14 | Claude Opus 4.7 (Adaptive Reasoning, Max Effort) | Anthropic | 73.6 | 51 tok/s | |
| #15 | Grok 4.5 (high) | SpaceXAI | 72.4 | 59 tok/s | |
| #16 | Muse Spark 1.2 (xhigh) | Meta | 72.2 | 0 tok/s | |
| #17 | Kimi K3 (low) | Kimi | 72.0 | 38 tok/s | |
| #18 | GPT-5.5 (high) | OpenAI | 71.6 | 72 tok/s | |
| #19 | Claude Sonnet 5 (Adaptive Reasoning, Max Effort) | Anthropic | 71.5 | 70 tok/s | |
| #20 | GPT-5.5 (medium) | OpenAI | 71.5 | 69 tok/s | |
| #21 | GPT-5.6 Luna (max) | OpenAI | 71.4 | 178 tok/s | |
| #22 | Muse Spark 1.1 (xhigh) | Meta | 71.3 | 213 tok/s | |
| #23 | GPT-5.4 (xhigh) | OpenAI | 71.1 | 125 tok/s | |
| #24 | GPT-5.6 Terra (xhigh) | OpenAI | 70.6 | 116 tok/s | |
| #25 | Gemini 3.5 Flash (high) | 70.1 | 168 tok/s | ||
| #26 | GPT-5.6 Sol (low) | OpenAI | 69.7 | 61 tok/s | |
| #27 | Gemini 3.6 Flash (high) | 69.2 | 188 tok/s | ||
| #28 | DeepSeek V4 Flash 0731 (Reasoning, Max Effort) | DeepSeek | 69.1 | 103 tok/s | |
| #29 | Gemini 3.1 Pro Preview | 68.8 | 126 tok/s | ||
| #30 | GLM-5.2 (max) | Z AI | 68.8 | 150 tok/s | |
| #31 | GPT-5.6 Luna (xhigh) | OpenAI | 68.6 | 164 tok/s | |
| #32 | GPT-5.6 Terra (high) | OpenAI | 67.1 | 109 tok/s | |
| #33 | Claude Opus 5 (Adaptive Reasoning, Low Effort) | Anthropic | 66.9 | 52 tok/s | |
| #34 | Claude Sonnet 5 (Non-reasoning, High Effort) | Anthropic | 66.4 | 62 tok/s | |
| #35 | Qwen3.7 Max | Alibaba | 66.0 | 202 tok/s | |
| #36 | GPT-5.6 Sol (Non-reasoning) | OpenAI | 65.1 | 60 tok/s | |
| #37 | GPT-5.6 Terra (medium) | OpenAI | 64.7 | 112 tok/s | |
| #38 | GPT-5.6 Luna (high) | OpenAI | 63.3 | 163 tok/s | |
| #39 | Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort) | Anthropic | 63.0 | 58 tok/s | |
| #40 | Motif 3 (Beta) | Motif Technologies | 62.0 | 0 tok/s | |
| #41 | Kimi K2.6 | Kimi | 61.8 | 39 tok/s | |
| #42 | GPT-5.5 (low) | OpenAI | 60.9 | 69 tok/s | |
| #43 | Kimi K2.7 Code | Kimi | 60.8 | 42 tok/s | |
| #44 | MiMo-V2.5-Pro | Xiaomi | 60.2 | 68 tok/s | |
| #45 | KAT Coder Pro V2 | KwaiKAT | 59.5 | 108 tok/s | |
| #46 | DeepSeek V4 Pro (Reasoning, Max Effort) | DeepSeek | 59.4 | 63 tok/s | |
| #47 | Nex-N2-Pro | Nex AGI | 59.1 | 129 tok/s | |
| #48 | Agnes 2.5 Pro Alpha | Sapiens AI | 58.8 | 129 tok/s | |
| #49 | Hy3 | Tencent | 58.8 | 65 tok/s | |
| #50 | DeepSeek V4 Pro (Reasoning, High Effort) | DeepSeek | 58.7 | 64 tok/s |