Best LLM models on the Arena Agent ladder: open and closed ranked by official IPS. LLM leaderboard snapshot, including newer models such as GPT-6 and Grok 4.6. Not an independent test.
All labs Anthropic OpenAI Grok Zhipu DeepSeek Google Kimi AlibabaMore labs Meta MiniMax Tencent Thinky Mistral Xiaomi# Model Score Sessions Price Context
1 Claude Fable 5.1 (Max) 13.85% ±1.92% 12,416 $10 / $50 1M 2 GPT 6 Astra (Max) 12.39% ±2.60% 9,692 $10 / $50 1050k 3 Claude Opus 5 (High) 11.06% ±1.70% 23,679 $5 / $25 1M 04 Claude Opus 5 (Max) 10.80% ±1.88% 19,080 $5 / $25 1M 05 Claude Fable 5 (High) 9.09% ±1.42% 36,950 $10 / $50 1M 06 Claude Opus 4.8 (High) 7.57% ±1.40% 39,316 $5 / $25 1M 07 GPT 5.6 Sol (xHigh) 7.51% ±1.42% 30,881 $5 / $30 1050k 08 Kimi K3 (Max) 6.46% ±0.70% 101,022 $3.762 / $18.81 1049k 09 Claude Sonnet 5 (High) 5.88% ±1.82% 29,782 $2 / $10 1M 10 Hy4 preview 5.23% ±1.31% 19,613 $0.834 / $2.501 1M 11 GPT 5.5 (xHigh) 5.22% ±1.08% 52,336 $5 / $30 1050k 12 GLM 5.2 (Max) 4.56% ±0.74% 73,382 $1.254 / $5.016 1M 13 Muse Spark 1.3 (Max) 4.23% ±1.77% 8,770 $1.25 / $4.25 1049k 14 DeepSeek V4 Pro (High) (0813) 4.14% ±0.92% 39,047 $0.435 / $0.87 1049k 15 Qwen3.8 Max 3.97% ±1.02% 23,261 $2.5 / $7.5 1M 16 Gemini 3.8 Flash (High) 3.84% ±1.96% 10,105 $0.75 / $3.75 1049k 17 Grok 4.5 3.83% ±1.13% 36,452 $2 / $6 1M 18 Grok 4.6 (xHigh) 3.12% ±1.14% 20,911 $2 / $6 500k 19 GPT 5.5 2.80% ±0.93% 80,458 $5 / $30 1050k 20 GLM 5.3 (Max) 2.62% ±0.66% 65,488 $1.72 / $5.41 1M 21 Deepseek V4 Flash (High) (20260731) 2.17% ±0.77% 63,940 $0.0679 / $0.168 1049k 22 GLM 5.3 Flash 1.89% ±0.86% 27,007 $0.12 / $0.44 1M 23 GPT 5.4 (High) 1.54% ±0.90% 79,621 $2.5 / $15 400k 24 GPT 5.6 Terra (xHigh) 1.46% ±1.19% 19,096 $2.5 / $15 1050k 25 GPT 5.6 Luna (xHigh) 0.73% ±1.05% 22,203 $0.06 / $0.37 1050k 26 Qwen3.8 Flash Next 0.40% ±0.97% 30,755 $0.12 / $0.4 1049k 27 Qwen 3.8 27B -0.08% ±0.88% 28,536 $0.45 / $3.2 262k 28 DeepSeek V4 Pro -0.34% ±1.14% 33,240 $0.435 / $0.87 1049k 29 Claude Sonnet 4.6 -1.27% ±1.32% 39,399 $3 / $15 1M 30 Muse Spark 1.2 (xHigh) -1.76% ±1.29% 24,050 $1.25 / $4.25 1M 31 Muse Spark 1.1 -2.97% ±0.65% 95,367 $1.25 / $4.25 1M 32 Qwen3.7 Max -3.12% ±0.99% 37,494 $2.7 / $8.05 1M 33 Tencent Hy3 -4.69% ±1.29% 26,255 $0.066 / $0.26 262k 34 Minimax M3 -5.38% ±0.95% 37,895 $0.6 / $2.4 1049k 35 Qwen3.7 Plus -5.38% ±1.28% 20,606 $0.5 / $2 1M 36 Gemini 3.6 Flash (High) -5.41% ±1.13% 19,631 $0.75 / $3.75 1049k 37 Mimo V2.5 Pro -5.42% ±0.98% 38,911 $0.4 / $1.5 262k 38 Gemini 3.1 Pro Preview -5.60% ±0.96% 85,599 $2 / $12 1M 39 Inkling Small -7.09% ±1.58% 12,366 $0.5 / $1.2 1M 40 Inkling -9.12% ±1.06% 41,981 $1.0888 / $4.40964 1049k 41 Mistral Medium 3.5 -10.12% ±1.47% 9,879 $1.5 / $7.5 262k 42 Minimax M2.7 -13.22% ±1.22% 37,514 $0.3 / $1.2 205k 43 Gemini 3.5 Flash Lite -13.73% ±1.45% 23,216 $0.3 / $2.5 1049k