Name: 智能体模型厂商排行榜（2026-07-06）- 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	14.07% ±1.56%

综合排名Overall Rank — 名次区间Rank Spread 1 ↔ 1 确认成功Confirmed Success ▲16.51%±2.73% 赞扬/抱怨Praise vs Complaint ▲30.91%±5.69% 可控性Steerability ▲12.65%±2.93% Bash 恢复Bash Recovery ▲8.99%±1.84% 工具幻觉Tool Hallucination ▼1.26%±0.15% 会话数Sessions 16,059
› 🥈	OpenAIOpenAI GPT 5.5 (xHigh) · Proprietary	8.39% ±0.94%

综合排名Overall Rank — 名次区间Rank Spread 2 ↔ 8 确认成功Confirmed Success ▲6.48%±1.82% 赞扬/抱怨Praise vs Complaint ▲13.32%±3.38% 可控性Steerability ▲6.41%±1.92% Bash 恢复Bash Recovery ▲14.46%±0.89% 工具幻觉Tool Hallucination ▼1.26%±0.15% 会话数Sessions 30,341
› 🥉	智谱 ZAIZ.ai GLM 5.2 (Max) · MIT · SiliconFlow	6.72% ±1.24%

综合排名Overall Rank — 名次区间Rank Spread 3 ↔ 11 确认成功Confirmed Success ▲9.41%±2.32% 赞扬/抱怨Praise vs Complaint ▲14.85%±4.50% 可控性Steerability ▲3.25%±2.46% Bash 恢复Bash Recovery ▲4.83%±1.24% 工具幻觉Tool Hallucination ▼1.26%±0.15% 会话数Sessions 27,177
› 4	月之暗面Moonshot kimi-k2.7-code · Modified MIT	0.74% ±3.33%

综合排名Overall Rank — 名次区间Rank Spread 11 ↔ 22 确认成功Confirmed Success ▲9.36%±5.66% 赞扬/抱怨Praise vs Complaint ▲2.99%±10.97% 可控性Steerability ▲11.45%±9.87% Bash 恢复Bash Recovery ▲1.52%±2.87% 工具幻觉Tool Hallucination ▼1.26%±0.15% 会话数Sessions 3,124
› 5	GoogleGoogle Gemini 3.1 Pro Preview · Proprietary	0.55% ±0.70%

综合排名Overall Rank — 名次区间Rank Spread 14 ↔ 19 确认成功Confirmed Success ▲0.22%±1.57% 赞扬/抱怨Praise vs Complaint ▲0.34%±2.21% 可控性Steerability ▲3.07%±1.30% Bash 恢复Bash Recovery ▲6.89%±1.23% 工具幻觉Tool Hallucination ▼1.21%±0.15% 会话数Sessions 55,576
› 6	DeepSeekDeepSeek DeepSeek V4 Flash · MIT · SiliconFlow	1.00% ±1.26%

综合排名Overall Rank — 名次区间Rank Spread 14 ↔ 22 确认成功Confirmed Success ▲4.48%±2.43% 赞扬/抱怨Praise vs Complaint ▲4.30%±4.46% 可控性Steerability ▲8.47%±2.42% Bash 恢复Bash Recovery ▲4.91%±1.79% 工具幻觉Tool Hallucination ▼0.39%±0.32% 会话数Sessions 40,093
› 7	MiniMaxMiniMax Minimax M3 · MiniMax Community License · Fireworks	2.45% ±1.09%

综合排名Overall Rank — 名次区间Rank Spread 15 ↔ 22 确认成功Confirmed Success ▲3.53%±2.66% 赞扬/抱怨Praise vs Complaint ▲7.43%±3.66% 可控性Steerability ▲4.94%±2.19% Bash 恢复Bash Recovery ▲2.38%±0.97% 工具幻觉Tool Hallucination ▼1.26%±0.15% 会话数Sessions 33,958
› 8	阿里巴巴Alibaba Qwen 3.6 Plus · Proprietary · Fireworks	4.77% ±1.04%

综合排名Overall Rank — 名次区间Rank Spread 20 ↔ 24 确认成功Confirmed Success ▲1.99%±2.20% 赞扬/抱怨Praise vs Complaint ▲5.39%±3.48% 可控性Steerability ▲10.30%±2.12% Bash 恢复Bash Recovery ▲3.95%±1.38% 工具幻觉Tool Hallucination ▼2.24%±0.47% 会话数Sessions 43,467
› 9	xAIxAI Grok 4.3 (High) · Proprietary	7.66% ±0.82%

综合排名Overall Rank — 名次区间Rank Spread 23 ↔ 29 确认成功Confirmed Success ▲8.84%±1.85% 赞扬/抱怨Praise vs Complaint ▲14.45%±2.16% 可控性Steerability ▲6.58%±1.48% Bash 恢复Bash Recovery ▲9.09%±2.11% 工具幻觉Tool Hallucination ▼0.68%±0.19% 会话数Sessions 35,625
› 10	NVIDIANvidia Nemotron 3 Ultra · OpenMDW-1.1	11.26% ±2.94%

综合排名Overall Rank — 名次区间Rank Spread 24 ↔ 31 确认成功Confirmed Success ▲8.88%±5.48% 赞扬/抱怨Praise vs Complaint ▲6.49%±9.33% 可控性Steerability ▲19.94%±5.80% Bash 恢复Bash Recovery ▲19.71%±6.92% 工具幻觉Tool Hallucination ▼1.28%±1.13% 会话数Sessions 8,294

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