Name: 智能体 AI 排行榜 - 最佳 Agent 模型对比 (2026-05-30 历史榜单)
Creator: LM Arena
License: https://creativecommons.org/licenses/by/4.0/
Keywords: AI 智能体模型排行榜,智能体 AI 排行榜,最佳 Agent 模型对比,AI 模型排行,智能体评测,净提升

排名Rank ⇕	模型Model ⇕
› 🥇	GPT 5.5 (High) OpenAI · Proprietary	10.66% ±1.60%

综合排名Overall Rank — 名次区间Rank Spread 1 ↔ 4 确认成功Confirmed Success ▲7.06%±2.70% 赞扬/抱怨Praise vs Complaint ▲14.95%±5.98% 可控性Steerability ▲12.03%±3.06% Bash 恢复Bash Recovery ▲17.73%±1.85% 工具幻觉Tool Hallucination ▼1.52%±0.21% 会话数Sessions 24,416
› 🥈	Claude Opus 4.7 (Thinking) Anthropic · Proprietary	9.47% ±1.50%

综合排名Overall Rank — 名次区间Rank Spread 1 ↔ 6 确认成功Confirmed Success ▲7.95%±2.71% 赞扬/抱怨Praise vs Complaint ▲12.18%±5.52% 可控性Steerability ▲9.04%±2.90% Bash 恢复Bash Recovery ▲16.69%±1.51% 工具幻觉Tool Hallucination ▼1.49%±0.21% 会话数Sessions 24,330
› 🥉	GPT 5.4 (High) OpenAI · Proprietary	8.92% ±1.68%

综合排名Overall Rank — 名次区间Rank Spread 1 ↔ 6 确认成功Confirmed Success ▲6.89%±2.85% 赞扬/抱怨Praise vs Complaint ▲9.72%±6.07% 可控性Steerability ▲10.12%±3.34% Bash 恢复Bash Recovery ▲16.34%±1.88% 工具幻觉Tool Hallucination ▼1.52%±0.21% 会话数Sessions 24,230
› 4	Claude Opus 4.6 Anthropic · Proprietary	8.14% ±1.46%

综合排名Overall Rank — 名次区间Rank Spread 1 ↔ 6 确认成功Confirmed Success ▲7.17%±2.67% 赞扬/抱怨Praise vs Complaint ▲8.69%±5.21% 可控性Steerability ▲7.06%±2.65% Bash 恢复Bash Recovery ▲16.26%±1.62% 工具幻觉Tool Hallucination ▼1.51%±0.21% 会话数Sessions 24,446
› 5	GPT 5.5 OpenAI · Proprietary	7.47% ±1.54%

综合排名Overall Rank — 名次区间Rank Spread 2 ↔ 7 确认成功Confirmed Success ▲2.97%±2.70% 赞扬/抱怨Praise vs Complaint ▲7.39%±5.48% 可控性Steerability ▲8.91%±2.92% Bash 恢复Bash Recovery ▲16.58%±1.82% 工具幻觉Tool Hallucination ▼1.52%±0.21% 会话数Sessions 24,623
› 6	Claude Opus 4.7 Anthropic · Proprietary	6.95% ±1.46%

综合排名Overall Rank — 名次区间Rank Spread 2 ↔ 7 确认成功Confirmed Success ▲5.46%±2.77% 赞扬/抱怨Praise vs Complaint ▲6.27%±5.30% 可控性Steerability ▲5.60%±2.73% Bash 恢复Bash Recovery ▲15.95%±1.53% 工具幻觉Tool Hallucination ▼1.48%±0.21% 会话数Sessions 24,499
› 7	Claude Sonnet 4.6 Anthropic · Proprietary	4.59% ±1.37%

综合排名Overall Rank — 名次区间Rank Spread 5 ↔ 8 确认成功Confirmed Success ▲1.22%±2.82% 赞扬/抱怨Praise vs Complaint ▲0.30%±4.74% 可控性Steerability ▲3.38%±2.70% Bash 恢复Bash Recovery ▲17.23%±1.75% 工具幻觉Tool Hallucination ▼1.43%±0.24% 会话数Sessions 24,361
› 8	GLM 5.1 智谱 ZAI · MIT	3.38% ±2.00%

综合排名Overall Rank — 名次区间Rank Spread 7 ↔ 10 确认成功Confirmed Success ▲4.63%±3.41% 赞扬/抱怨Praise vs Complaint ▲3.79%±7.15% 可控性Steerability ▲3.41%±4.15% Bash 恢复Bash Recovery ▲10.37%±2.22% 工具幻觉Tool Hallucination ▼1.52%±0.21% 会话数Sessions 19,560
› 9	Gemini 3.1 Pro Preview Google · Proprietary	1.38% ±1.45%

综合排名Overall Rank — 名次区间Rank Spread 8 ↔ 11 确认成功Confirmed Success ▲0.64%±2.79% 赞扬/抱怨Praise vs Complaint ▲1.52%±4.57% 可控性Steerability ▲4.33%±2.65% Bash 恢复Bash Recovery ▲0.95%±2.78% 工具幻觉Tool Hallucination ▼1.34%±0.32% 会话数Sessions 24,293
› 10	Gemini 3.5 Flash Google · Proprietary	0.40% ±1.82%

综合排名Overall Rank — 名次区间Rank Spread 8 ↔ 12 确认成功Confirmed Success ▲2.46%±3.84% 赞扬/抱怨Praise vs Complaint ▲0.08%±5.88% 可控性Steerability ▲1.17%±3.42% Bash 恢复Bash Recovery ▲4.08%±3.02% 工具幻觉Tool Hallucination ▼1.49%±0.21% 会话数Sessions 17,470
› 11	Kimi K2.6 月之暗面 · Modified MIT	0.56% ±1.64%

综合排名Overall Rank — 名次区间Rank Spread 9 ↔ 13 确认成功Confirmed Success ▲0.88%±3.39% 赞扬/抱怨Praise vs Complaint ▲3.70%±5.68% 可控性Steerability ▲8.44%±3.42% Bash 恢复Bash Recovery ▲8.68%±2.30% 工具幻觉Tool Hallucination ▼1.52%±0.21% 会话数Sessions 21,043
› 12	DeepSeek V4 Pro DeepSeek · MIT	1.88% ±1.79%

综合排名Overall Rank — 名次区间Rank Spread 10 ↔ 14 确认成功Confirmed Success ▲1.57%±3.43% 赞扬/抱怨Praise vs Complaint ▲1.44%±6.27% 可控性Steerability ▲1.26%±3.62% Bash 恢复Bash Recovery ▲2.78%±2.31% 工具幻觉Tool Hallucination ▼5.48%±1.25% 会话数Sessions 19,813
› 13	Qwen 3.6 Plus 阿里巴巴 · Proprietary	3.40% ±1.91%

综合排名Overall Rank — 名次区间Rank Spread 11 ↔ 14 确认成功Confirmed Success ▲0.48%±3.61% 赞扬/抱怨Praise vs Complaint ▲5.74%±6.33% 可控性Steerability ▲10.84%±4.14% Bash 恢复Bash Recovery ▲2.37%±2.99% 工具幻觉Tool Hallucination ▼2.32%±0.95% 会话数Sessions 19,286
› 14	DeepSeek V4 Flash DeepSeek · MIT	5.10% ±1.77%

综合排名Overall Rank — 名次区间Rank Spread 12 ↔ 15 确认成功Confirmed Success ▲0.00%±3.58% 赞扬/抱怨Praise vs Complaint ▲8.65%±5.74% 可控性Steerability ▲15.29%±3.91% Bash 恢复Bash Recovery ▲1.68%±2.23% 工具幻觉Tool Hallucination ▼3.22%±0.98% 会话数Sessions 19,707
› 15	Minimax M2.7 MiniMax · Modified MIT	8.52% ±1.76%

综合排名Overall Rank — 名次区间Rank Spread 14 ↔ 17 确认成功Confirmed Success ▲8.00%±3.63% 赞扬/抱怨Praise vs Complaint ▲15.73%±5.28% 可控性Steerability ▲17.56%±3.60% Bash 恢复Bash Recovery ▲2.84%±4.08% 工具幻觉Tool Hallucination ▼1.52%±0.21% 会话数Sessions 19,802
› 16	Gemini 3 Flash Google · Proprietary	9.22% ±1.57%

综合排名Overall Rank — 名次区间Rank Spread 15 ↔ 17 确认成功Confirmed Success ▲15.64%±3.02% 赞扬/抱怨Praise vs Complaint ▲15.63%±3.82% 可控性Steerability ▲5.81%±2.78% Bash 恢复Bash Recovery ▲7.77%±4.45% 工具幻觉Tool Hallucination ▼1.27%±1.99% 会话数Sessions 24,298
› 17	Gemma 4 31B Google · Apache 2.0	14.63% ±4.98%

综合排名Overall Rank — 名次区间Rank Spread 15 ↔ 17 确认成功Confirmed Success ▲4.16%±5.75% 赞扬/抱怨Praise vs Complaint ▲7.65%±8.08% 可控性Steerability ▲6.86%±5.48% Bash 恢复Bash Recovery ▲21.86%±15.92% 工具幻觉Tool Hallucination ▼32.64%±15.97% 会话数Sessions 13,482
› 18	Grok 4.3 xAI · Proprietary	25.14% ±2.26%

综合排名Overall Rank — 名次区间Rank Spread 18 ↔ 18 确认成功Confirmed Success ▲16.12%±3.03% 赞扬/抱怨Praise vs Complaint ▲17.51%±3.70% 可控性Steerability ▲3.90%±2.78% Bash 恢复Bash Recovery ▲89.43%±9.34% 工具幻觉Tool Hallucination ▼1.24%±0.35% 会话数Sessions 23,473

没有找到相关模型No matching models found

AI 智能体模型排行榜Agent Models Leaderboard 🏆 综合榜单