Name: 智能体模型厂商排行榜（2026-06-19）- 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	14.05% ±1.53%

综合排名Overall Rank — 名次区间Rank Spread 1 ↔ 1 确认成功Confirmed Success ▲16.27%±2.58% 赞扬/抱怨Praise vs Complaint ▲29.86%±5.65% 可控性Steerability ▲12.34%±2.93% Bash 恢复Bash Recovery ▲10.04%±1.81% 工具幻觉Tool Hallucination ▼1.75%±0.21% 会话数Sessions 16,232
› 🥈	OpenAIOpenAI GPT 5.5 (xHigh) · Proprietary	8.13% ±1.41%

综合排名Overall Rank — 名次区间Rank Spread 2 ↔ 9 确认成功Confirmed Success ▲5.13%±2.62% 赞扬/抱怨Praise vs Complaint ▲14.82%±5.22% 可控性Steerability ▲3.95%±2.75% Bash 恢复Bash Recovery ▲15.00%±1.26% 工具幻觉Tool Hallucination ▼1.75%±0.21% 会话数Sessions 13,752
› 🥉	智谱 ZAIZ.ai GLM 5.2 (Max) · MIT · SiliconFlow	4.51% ±1.77%

综合排名Overall Rank — 名次区间Rank Spread 7 ↔ 13 确认成功Confirmed Success ▲9.96%±3.23% 赞扬/抱怨Praise vs Complaint ▲12.69%±6.49% 可控性Steerability ▲5.45%±3.38% Bash 恢复Bash Recovery ▲3.60%±1.98% 工具幻觉Tool Hallucination ▼1.75%±0.21% 会话数Sessions 13,368
› 4	GoogleGoogle Gemini 3.5 Flash · Proprietary	0.03% ±0.84%

综合排名Overall Rank — 名次区间Rank Spread 14 ↔ 19 确认成功Confirmed Success ▲1.06%±1.85% 赞扬/抱怨Praise vs Complaint ▲1.71%±2.74% 可控性Steerability ▲1.11%±1.60% Bash 恢复Bash Recovery ▲2.68%±1.23% 工具幻觉Tool Hallucination ▼1.36%±0.25% 会话数Sessions 33,308
› 5	DeepSeekDeepSeek DeepSeek V4 Pro · MIT · SiliconFlow	0.76% ±1.25%

综合排名Overall Rank — 名次区间Rank Spread 14 ↔ 20 确认成功Confirmed Success ▲0.75%±2.76% 赞扬/抱怨Praise vs Complaint ▲1.84%±4.32% 可控性Steerability ▲3.64%±2.57% Bash 恢复Bash Recovery ▲2.27%±1.15% 工具幻觉Tool Hallucination ▼0.13%±0.34% 会话数Sessions 28,713
› 6	月之暗面Moonshot Kimi K2.6 · Modified MIT · Fireworks	1.01% ±0.88%

综合排名Overall Rank — 名次区间Rank Spread 14 ↔ 20 确认成功Confirmed Success ▲0.43%±1.88% 赞扬/抱怨Praise vs Complaint ▲2.97%±2.91% 可控性Steerability ▲3.51%±1.76% Bash 恢复Bash Recovery ▲0.11%±1.35% 工具幻觉Tool Hallucination ▼1.75%±0.21% 会话数Sessions 36,856
› 7	MiniMaxMiniMax Minimax M3 · Proprietary · Fireworks	2.04% ±1.23%

综合排名Overall Rank — 名次区间Rank Spread 15 ↔ 21 确认成功Confirmed Success ▲1.17%±2.77% 赞扬/抱怨Praise vs Complaint ▲6.97%±4.07% 可控性Steerability ▲7.42%±2.62% Bash 恢复Bash Recovery ▲3.61%±1.15% 工具幻觉Tool Hallucination ▼1.75%±0.21% 会话数Sessions 15,881
› 8	阿里巴巴Alibaba Qwen 3.6 Plus · Proprietary · Fireworks	4.12% ±0.97%

综合排名Overall Rank — 名次区间Rank Spread 20 ↔ 22 确认成功Confirmed Success ▲0.56%±1.97% 赞扬/抱怨Praise vs Complaint ▲6.90%±3.21% 可控性Steerability ▲9.75%±2.04% Bash 恢复Bash Recovery ▲1.62%±1.37% 工具幻觉Tool Hallucination ▼1.76%±0.52% 会话数Sessions 34,824
› 9	xAIxAI Grok Build 0.1 · Proprietary	6.26% ±0.95%

综合排名Overall Rank — 名次区间Rank Spread 22 ↔ 25 确认成功Confirmed Success ▲6.85%±2.22% 赞扬/抱怨Praise vs Complaint ▲11.34%±2.99% 可控性Steerability ▲9.37%±2.00% Bash 恢复Bash Recovery ▲2.00%±1.42% 工具幻觉Tool Hallucination ▼1.74%±0.38% 会话数Sessions 30,100
› 10	NVIDIANvidia Nemotron 3 Ultra · OpenMDW-1.1	7.36% ±3.74%

综合排名Overall Rank — 名次区间Rank Spread 21 ↔ 27 确认成功Confirmed Success ▲5.54%±6.82% 赞扬/抱怨Praise vs Complaint ▲2.30%±13.04% 可控性Steerability ▲19.65%±7.29% Bash 恢复Bash Recovery ▲10.21%±6.70% 工具幻觉Tool Hallucination ▼0.90%±0.99% 会话数Sessions 4,676

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