Name: 智能体模型厂商排行榜（2026-07-07）- 最佳 Agent 模型对比 - 厂商聚合排名
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 ⇕	厂商Lab ⇕
› 🥇	AnthropicAnthropic Claude Fable 5 (High) · Proprietary	14.06% ±1.56%

综合排名Overall Rank — 名次区间Rank Spread 1 ↔ 1 确认成功Confirmed Success ▲16.58%±2.73% 赞扬/抱怨Praise vs Complaint ▲30.97%±5.69% 可控性Steerability ▲12.63%±2.93% Bash 恢复Bash Recovery ▲8.89%±1.85% 工具幻觉Tool Hallucination ▼1.24%±0.15% 会话数Sessions 16,059
› 🥈	OpenAIOpenAI GPT 5.5 (xHigh) · Proprietary	8.52% ±0.92%

综合排名Overall Rank — 名次区间Rank Spread 2 ↔ 8 确认成功Confirmed Success ▲6.91%±1.77% 赞扬/抱怨Praise vs Complaint ▲13.02%±3.31% 可控性Steerability ▲6.93%±1.88% Bash 恢复Bash Recovery ▲14.48%±0.88% 工具幻觉Tool Hallucination ▼1.24%±0.15% 会话数Sessions 31,540
› 🥉	智谱 ZAIZ.ai GLM 5.2 (Max) · MIT · SiliconFlow	6.62% ±1.20%

综合排名Overall Rank — 名次区间Rank Spread 3 ↔ 12 确认成功Confirmed Success ▲8.88%±2.26% 赞扬/抱怨Praise vs Complaint ▲14.70%±4.36% 可控性Steerability ▲3.41%±2.38% Bash 恢复Bash Recovery ▲4.88%±1.19% 工具幻觉Tool Hallucination ▼1.24%±0.15% 会话数Sessions 28,435
› 4	GoogleGoogle Gemini 3.1 Pro Preview · Proprietary	0.41% ±0.69%

综合排名Overall Rank — 名次区间Rank Spread 15 ↔ 18 确认成功Confirmed Success ▲0.51%±1.55% 赞扬/抱怨Praise vs Complaint ▲0.15%±2.20% 可控性Steerability ▲3.35%±1.29% Bash 恢复Bash Recovery ▲6.96%±1.21% 工具幻觉Tool Hallucination ▼1.20%±0.15% 会话数Sessions 56,839
› 5	月之暗面Moonshot Kimi K2.7 Code · Modified MIT · Fireworks	1.06% ±1.35%

综合排名Overall Rank — 名次区间Rank Spread 15 ↔ 20 确认成功Confirmed Success ▲0.24%±2.93% 赞扬/抱怨Praise vs Complaint ▲2.71%±4.50% 可控性Steerability ▲2.12%±2.69% Bash 恢复Bash Recovery ▲1.46%±2.27% 工具幻觉Tool Hallucination ▼1.24%±0.15% 会话数Sessions 28,613
› 6	DeepSeekDeepSeek DeepSeek V4 Flash · MIT · SiliconFlow	1.19% ±1.28%

综合排名Overall Rank — 名次区间Rank Spread 15 ↔ 20 确认成功Confirmed Success ▲4.61%±2.50% 赞扬/抱怨Praise vs Complaint ▲3.82%±4.51% 可控性Steerability ▲8.77%±2.48% Bash 恢复Bash Recovery ▲5.21%±1.81% 工具幻觉Tool Hallucination ▼0.39%±0.31% 会话数Sessions 40,268
› 7	MiniMaxMiniMax Minimax M3 · MiniMax Community License · Fireworks	2.26% ±1.11%

综合排名Overall Rank — 名次区间Rank Spread 16 ↔ 20 确认成功Confirmed Success ▲2.96%±2.70% 赞扬/抱怨Praise vs Complaint ▲7.22%±3.73% 可控性Steerability ▲4.98%±2.24% Bash 恢复Bash Recovery ▲2.63%±0.96% 工具幻觉Tool Hallucination ▼1.24%±0.15% 会话数Sessions 34,335
› 8	阿里巴巴Alibaba Qwen 3.6 Plus · Proprietary · Fireworks	4.73% ±1.06%

综合排名Overall Rank — 名次区间Rank Spread 20 ↔ 22 确认成功Confirmed Success ▲1.83%±2.22% 赞扬/抱怨Praise vs Complaint ▲5.22%±3.50% 可控性Steerability ▲10.07%±2.16% Bash 恢复Bash Recovery ▲4.26%±1.40% 工具幻觉Tool Hallucination ▼2.27%±0.47% 会话数Sessions 43,467
› 9	xAIxAI Grok 4.3 (High) · Proprietary	7.64% ±0.80%

综合排名Overall Rank — 名次区间Rank Spread 22 ↔ 26 确认成功Confirmed Success ▲9.06%±1.82% 赞扬/抱怨Praise vs Complaint ▲14.10%±2.12% 可控性Steerability ▲6.67%±1.46% Bash 恢复Bash Recovery ▲9.08%±2.08% 工具幻觉Tool Hallucination ▼0.70%±0.18% 会话数Sessions 36,897
› 10	NVIDIANvidia Nemotron 3 Ultra · OpenMDW-1.1	11.75% ±2.88%

综合排名Overall Rank — 名次区间Rank Spread 24 ↔ 28 确认成功Confirmed Success ▲10.10%±5.46% 赞扬/抱怨Praise vs Complaint ▲6.89%±9.08% 可控性Steerability ▲21.04%±5.64% Bash 恢复Bash Recovery ▲19.53%±6.77% 工具幻觉Tool Hallucination ▼1.17%±1.07% 会话数Sessions 8,556

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