Name: 智能体模型厂商排行榜（2026-06-15）- 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	14.17% ±1.54%

综合排名Overall Rank — 名次区间Rank Spread 1 ↔ 1 确认成功Confirmed Success ▲16.49%±2.54% 赞扬/抱怨Praise vs Complaint ▲29.62%±5.72% 可控性Steerability ▲13.32%±2.94% Bash 恢复Bash Recovery ▲9.52%±1.85% 工具幻觉Tool Hallucination ▼1.88%±0.23% 会话数Sessions 16,246
› 🥈	OpenAIOpenAI GPT 5.5 (xHigh) · Proprietary	9.36% ±1.78%

综合排名Overall Rank — 名次区间Rank Spread 2 ↔ 8 确认成功Confirmed Success ▲5.52%±3.20% 赞扬/抱怨Praise vs Complaint ▲19.55%±6.67% 可控性Steerability ▲5.35%±3.54% Bash 恢复Bash Recovery ▲14.48%±1.27% 工具幻觉Tool Hallucination ▼1.88%±0.23% 会话数Sessions 10,493
› 🥉	智谱 ZAIZ.ai GLM 5.1 · MIT · SiliconFlow	2.72% ±1.15%

综合排名Overall Rank — 名次区间Rank Spread 10 ↔ 13 确认成功Confirmed Success ▲3.66%±2.36% 赞扬/抱怨Praise vs Complaint ▲1.78%±4.02% 可控性Steerability ▲1.04%±2.35% Bash 恢复Bash Recovery ▲5.25%±1.13% 工具幻觉Tool Hallucination ▼1.88%±0.23% 会话数Sessions 32,039
› 4	DeepSeekDeepSeek DeepSeek V4 Pro · MIT · SiliconFlow	0.30% ±1.39%

综合排名Overall Rank — 名次区间Rank Spread 12 ↔ 17 确认成功Confirmed Success ▲0.22%±2.99% 赞扬/抱怨Praise vs Complaint ▲0.42%±4.81% 可控性Steerability ▲2.60%±2.97% Bash 恢复Bash Recovery ▲2.71%±1.34% 工具幻觉Tool Hallucination ▼0.73%±0.33% 会话数Sessions 27,932
› 5	GoogleGoogle Gemini 3.5 Flash · Proprietary	0.08% ±1.00%

综合排名Overall Rank — 名次区间Rank Spread 13 ↔ 17 确认成功Confirmed Success ▲2.24%±2.26% 赞扬/抱怨Praise vs Complaint ▲2.59%±3.26% 可控性Steerability ▲0.14%±1.90% Bash 恢复Bash Recovery ▲2.87%±1.28% 工具幻觉Tool Hallucination ▼1.70%±0.27% 会话数Sessions 30,036
› 6	月之暗面Moonshot Kimi K2.6 · Modified MIT · Fireworks	0.38% ±1.10%

综合排名Overall Rank — 名次区间Rank Spread 13 ↔ 17 确认成功Confirmed Success ▲0.06%±2.29% 赞扬/抱怨Praise vs Complaint ▲0.93%±3.69% 可控性Steerability ▲2.52%±2.20% Bash 恢复Bash Recovery ▲0.38%±1.72% 工具幻觉Tool Hallucination ▼1.88%±0.23% 会话数Sessions 33,594
› 7	MiniMaxMiniMax Minimax M3 · Proprietary · Fireworks	3.43% ±1.83%

综合排名Overall Rank — 名次区间Rank Spread 16 ↔ 21 确认成功Confirmed Success ▲2.27%±4.27% 赞扬/抱怨Praise vs Complaint ▲10.37%±5.82% 可控性Steerability ▲9.35%±4.14% Bash 恢复Bash Recovery ▲2.96%±1.54% 工具幻觉Tool Hallucination ▼1.88%±0.23% 会话数Sessions 9,247
› 8	阿里巴巴Alibaba Qwen 3.6 Plus · Proprietary · Fireworks	4.26% ±1.20%

综合排名Overall Rank — 名次区间Rank Spread 18 ↔ 21 确认成功Confirmed Success ▲2.62%±2.44% 赞扬/抱怨Praise vs Complaint ▲5.99%±4.06% 可控性Steerability ▲9.12%±2.48% Bash 恢复Bash Recovery ▲2.28%±1.64% 工具幻觉Tool Hallucination ▼1.31%±0.54% 会话数Sessions 31,638
› 9	xAIxAI Grok Build 0.1 · Proprietary	6.25% ±1.11%

综合排名Overall Rank — 名次区间Rank Spread 18 ↔ 23 确认成功Confirmed Success ▲7.46%±2.58% 赞扬/抱怨Praise vs Complaint ▲11.17%±3.58% 可控性Steerability ▲8.68%±2.25% Bash 恢复Bash Recovery ▲1.38%±1.63% 工具幻觉Tool Hallucination ▼2.59%±0.49% 会话数Sessions 26,916
› 10	NVIDIANvidia Nemotron 3 Ultra · OpenMDW-1.1	8.89% ±4.15%

综合排名Overall Rank — 名次区间Rank Spread 18 ↔ 25 确认成功Confirmed Success ▲4.47%±7.71% 赞扬/抱怨Praise vs Complaint ▲6.28%±14.19% 可控性Steerability ▲22.74%±8.38% Bash 恢复Bash Recovery ▲12.82%±7.89% 工具幻觉Tool Hallucination ▼1.88%±0.23% 会话数Sessions 3,851

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