Name: 智能体模型厂商排行榜（2026-06-30）- 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	13.34% ±1.55%

综合排名Overall Rank — 名次区间Rank Spread 1 ↔ 1 确认成功Confirmed Success ▲16.12%±2.68% 赞扬/抱怨Praise vs Complaint ▲30.63%±5.69% 可控性Steerability ▲9.21%±2.92% Bash 恢复Bash Recovery ▲9.40%±1.83% 工具幻觉Tool Hallucination ▼1.31%±0.13% 会话数Sessions 16,082
› 🥈	OpenAIOpenAI GPT 5.5 (xHigh) · Proprietary	8.21% ±1.02%

综合排名Overall Rank — 名次区间Rank Spread 2 ↔ 9 确认成功Confirmed Success ▲5.84%±1.96% 赞扬/抱怨Praise vs Complaint ▲13.63%±3.72% 可控性Steerability ▲5.78%±2.02% Bash 恢复Bash Recovery ▲14.50%±1.02% 工具幻觉Tool Hallucination ▼1.31%±0.13% 会话数Sessions 24,393
› 🥉	智谱 ZAIZ.ai GLM 5.2 (Max) · MIT · SiliconFlow	6.93% ±1.40%

综合排名Overall Rank — 名次区间Rank Spread 2 ↔ 10 确认成功Confirmed Success ▲9.13%±2.66% 赞扬/抱怨Praise vs Complaint ▲15.45%±5.27% 可控性Steerability ▲3.58%±2.59% Bash 恢复Bash Recovery ▲5.19%±1.27% 工具幻觉Tool Hallucination ▼1.31%±0.13% 会话数Sessions 21,946
› 4	月之暗面Moonshot Kimi K2.7 Code · Modified MIT · Fireworks	0.77% ±1.24%

综合排名Overall Rank — 名次区间Rank Spread 14 ↔ 20 确认成功Confirmed Success ▲0.82%±2.62% 赞扬/抱怨Praise vs Complaint ▲2.63%±4.26% 可控性Steerability ▲1.65%±2.52% Bash 恢复Bash Recovery ▲1.72%±2.19% 工具幻觉Tool Hallucination ▼1.31%±0.13% 会话数Sessions 26,189
› 5	GoogleGoogle Gemini 3.1 Pro Preview · Proprietary	1.09% ±0.70%

综合排名Overall Rank — 名次区间Rank Spread 14 ↔ 20 确认成功Confirmed Success ▲0.16%±1.58% 赞扬/抱怨Praise vs Complaint ▲1.88%±2.23% 可控性Steerability ▲1.37%±1.31% Bash 恢复Bash Recovery ▲6.34%±1.27% 工具幻觉Tool Hallucination ▼1.26%±0.14% 会话数Sessions 49,868
› 6	DeepSeekDeepSeek DeepSeek V4 Flash · MIT · SiliconFlow	1.57% ±1.08%

综合排名Overall Rank — 名次区间Rank Spread 14 ↔ 20 确认成功Confirmed Success ▲4.27%±2.02% 赞扬/抱怨Praise vs Complaint ▲1.61%±3.83% 可控性Steerability ▲9.19%±2.19% Bash 恢复Bash Recovery ▲4.02%±1.54% 工具幻觉Tool Hallucination ▼0.50%±0.32% 会话数Sessions 38,964
› 7	MiniMaxMiniMax Minimax M3 · Proprietary · Fireworks	2.28% ±1.01%

综合排名Overall Rank — 名次区间Rank Spread 14 ↔ 21 确认成功Confirmed Success ▲2.14%±2.34% 赞扬/抱怨Praise vs Complaint ▲7.20%±3.50% 可控性Steerability ▲5.11%±1.98% Bash 恢复Bash Recovery ▲1.75%±1.00% 工具幻觉Tool Hallucination ▼1.31%±0.13% 会话数Sessions 31,887
› 8	阿里巴巴Alibaba Qwen 3.6 Plus · Proprietary · Fireworks	4.20% ±0.99%

综合排名Overall Rank — 名次区间Rank Spread 19 ↔ 21 确认成功Confirmed Success ▲1.26%±1.97% 赞扬/抱怨Praise vs Complaint ▲3.80%±3.47% 可控性Steerability ▲10.58%±1.94% Bash 恢复Bash Recovery ▲3.03%±1.25% 工具幻觉Tool Hallucination ▼2.36%±0.47% 会话数Sessions 41,967
› 9	xAIxAI Grok 4.3 (High) · Proprietary	7.19% ±0.84%

综合排名Overall Rank — 名次区间Rank Spread 22 ↔ 26 确认成功Confirmed Success ▲9.41%±1.95% 赞扬/抱怨Praise vs Complaint ▲14.44%±2.35% 可控性Steerability ▲6.79%±1.53% Bash 恢复Bash Recovery ▲5.73%±1.88% 工具幻觉Tool Hallucination ▼0.44%±0.22% 会话数Sessions 29,641
› 10	NVIDIANvidia Nemotron 3 Ultra · OpenMDW-1.1	8.87% ±3.19%

综合排名Overall Rank — 名次区间Rank Spread 22 ↔ 27 确认成功Confirmed Success ▲5.53%±5.67% 赞扬/抱怨Praise vs Complaint ▲3.94%±10.80% 可控性Steerability ▲17.20%±6.19% Bash 恢复Bash Recovery ▲15.30%±6.34% 工具幻觉Tool Hallucination ▼2.37%±1.65% 会话数Sessions 7,048

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