LLM API 服务商排行榜——比较 500 多个 AI 模型接入点
比较 OpenAI、Google、DeepSeek 等 API 服务商在 500 多个 AI 模型接入点上的表现,涵盖价格、输出速度、延迟、上下文窗口等关键性能指标。
如需了解包括方法论在内的更多详情,请参阅我们的常见问题。
进一步分析 | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
Claude Opus 5 (max) | 1M | 专有 | 63 | $2.34 | 55 | 57.75 | 66.87 | -- | ||
Claude Opus 5 (max) | 1M | 专有 | 63 | $2.34 | 53 | 53.48 | 62.96 | -- | ||
Claude Opus 5 (max) | 1M | 专有 | 63 | $3.31 | 53 | 51.14 | 60.60 | -- | ||
Claude Opus 5 (xhigh) | 1M | 专有 | 63 | $1.80 | 54 | 25.75 | 35.05 | -- | ||
Claude Opus 5 (xhigh) | 1M | 专有 | 63 | $2.51 | 54 | 24.20 | 33.41 | -- | ||
Claude Opus 5 (xhigh) | 1M | 专有 | 63 | $1.80 | 54 | 33.07 | 42.35 | -- | ||
Claude Fable 5 (with fallback) | 1M | 专有 | 62 | $3.14 | 63 | 98.20 | 106.16 | -- | ||
Claude Fable 5 (with fallback) | 1M | 专有 | 62 | $3.32 | 64 | 96.99 | 104.75 | -- | ||
Claude Fable 5 (with fallback) | 1M | 专有 | 62 | $10.16 | 63 | 122.27 | 130.15 | -- | ||
Claude Opus 5 (high) | 1M | 专有 | 61 | $1.23 | 54 | 13.43 | 22.62 | -- | ||
Claude Opus 5 (high) | 1M | 专有 | 61 | $1.67 | 55 | 12.66 | 21.83 | -- | ||
Claude Opus 5 (high) | 1M | 专有 | 61 | $1.23 | 53 | 18.71 | 28.21 | -- | ||
GPT-5.6 Sol (max) | 1M | 专有 | 61 | $1.66 | 106 | 83.59 | 88.29 | -- | ||
GPT-5.6 Sol (max) | 1M | 专有 | 61 | $1.23 | 62 | 176.57 | 184.67 | -- | ||
Grok 4.6 (high) | 500k | 专有 | 61 | $0.84 | 66 | 32.30 | 39.89 | -- | ||
Kimi K3 (max) | 1.05M | 开放 | 60 | $0.84 | 39 | 3.64 | 66.95 | 50.65 | ||
Kimi K3 (max) | 1.05M | 开放 | 60 | $0.89 | 48 | 1.26 | 53.76 | 42.00 | ||
Kimi K3 (max) (FAST) | 1.05M | 开放 | 60 | $1.34 | 84 | 1.26 | 30.89 | 23.71 | ||
Kimi K3 (max) | 1.05M | 开放 | 60 | $1.24 | 141 | 2.11 | 19.82 | 14.16 | ||
Kimi K3 (max) | 1M | 开放 | 60 | $1.43 | 59 | 1.42 | 44.03 | 34.08 | ||
Kimi K3 (max) | 1.05M | 开放 | 60 | $1.01 | 37 | 1.24 | 69.18 | 54.36 | ||
Kimi K3 (max) | 1.05M | 开放 | 60 | $1.26 | 109 | 1.19 | 24.23 | 18.43 | ||
Kimi K3 (max) | 1.05M | 开放 | 60 | $0.78 | 147 | 2.02 | 18.98 | 13.56 | ||
Kimi K3 (max) | 1.05M | 开放 | 60 | $0.66 | 112 | 1.41 | 23.77 | 17.89 | ||
Kimi K3 (max) | 1.05M | 开放 | 60 | $3.11 | 130 | 1.60 | 20.88 | 15.42 | ||
Kimi K3 (max) | 205k | 开放 | 60 | $0.79 | 94 | 1.31 | 27.79 | 21.18 | ||
Kimi K3 (max) (FAST) | 1M | 开放 | 60 | $1.27 | 57 | 2.22 | 45.77 | 34.84 | ||
GPT-5.6 Sol (xhigh) | 1M | 专有 | 59 | $0.81 | 61 | 42.36 | 50.60 | -- | ||
Claude Opus 5 (medium) | 1M | 专有 | 59 | $0.72 | 54 | 5.76 | 15.10 | -- | ||
Claude Opus 5 (medium) | 1M | 专有 | 59 | $0.72 | 53 | 9.48 | 18.90 | -- | ||
Claude Opus 5 (medium) | 1M | 专有 | 59 | $0.95 | 52 | 5.04 | 14.58 | -- | ||
Qwen3.8 Max | 1M | 专有 | 58 | $1.13 | 47 | 2.60 | 55.64 | 42.43 | ||
Qwen3.8 2.4T A95B | 984k | 开放 | 58 | $1.09 | 52 | 2.75 | 51.26 | 38.81 | ||
GPT-5.6 Sol (high) | 1M | 专有 | 57 | $0.55 | 56 | 12.34 | 21.29 | -- | ||
Muse Spark 1.2 (xhigh) | 1.05M | 专有 | 57 | $0.40 | -- | -- | -- | -- | ||
GPT-5.6 Terra (max) | 1M | 专有 | 57 | $0.51 | 115 | 197.30 | 201.63 | -- | ||
GPT-5.6 Terra (max) | 1M | 专有 | 57 | $0.97 | 151 | 130.56 | 133.87 | -- | ||
Gemini 3.7 Flash (high) AI Studio | 1M | 专有 | 56 | $0.40 | 340 | 9.83 | 11.30 | -- | ||
Grok 4.5 (high) | 500k | 专有 | 56 | $0.36 | 60 | 8.14 | 16.41 | -- | ||
GPT-5.6 Sol (medium) | 1M | 专有 | 56 | $0.37 | 62 | 5.00 | 13.05 | -- | ||
Claude Sonnet 5 (max) | 1M | 专有 | 55 | $2.22 | 74 | 214.86 | 221.66 | -- | ||
Claude Sonnet 5 (max) | 1M | 专有 | 55 | $1.64 | 80 | 161.21 | 167.50 | -- | ||
Claude Sonnet 5 (max) | 1M | 专有 | 55 | $1.72 | 66 | 186.44 | 194.02 | -- | ||
Claude Sonnet 5 (max) | 1M | 专有 | 55 | $1.96 | 74 | 203.40 | 210.15 | -- | ||
Gemini 3.7 Flash (medium) AI Studio | 1M | 专有 | 53 | $0.26 | 274 | 4.22 | 6.05 | -- | ||
DeepSeek V4 Pro 0813 (max) | 1.05M | 开放 | 53 | $0.25 | 81 | 1.85 | 32.79 | 24.75 | ||
GPT-5.6 Terra (xhigh) | 1M | 专有 | 53 | $0.31 | 94 | 25.72 | 31.01 | -- | ||
GLM-5.2 (max) | 205k | 开放 | 53 | $0.36 | 165 | 6.58 | 21.73 | 12.12 | ||
GLM-5.2 (max) | 1M | 开放 | 53 | $0.42 | 184 | 1.37 | 14.92 | 10.84 | ||
GLM-5.2 (max) (NVFP4) | 1M | 开放 | 53 | $0.48 | 312 | 0.88 | 8.90 | 6.42 | ||
GLM-5.2 (max) (FAST) | 524k | 开放 | 53 | $0.54 | 206 | 1.90 | 14.01 | 9.69 | ||
GLM-5.2 (max) (FP8) | 1.05M | 开放 | 53 | $0.42 | 69 | 2.24 | 38.35 | 28.89 | ||
GLM-5.2 (max) | 261k | 开放 | 53 | $0.46 | 107 | 2.01 | 25.46 | 18.76 | ||
GLM-5.2 (max) (NVFP4) | 1M | 开放 | 53 | -- | 120 | 1.37 | 22.22 | 16.68 | ||
GLM-5.2 (max) | 262k | 开放 | 53 | $0.38 | 187 | 1.42 | 14.76 | 10.67 | ||
GLM-5.2 (max) (FP4) | 1.05M | 开放 | 53 | $0.25 | 36 | 1.41 | 71.73 | 56.26 | ||
GLM-5.2 (max) (FP4) | 432k | 开放 | 53 | $1.06 | 215 | 1.12 | 12.77 | 9.32 | ||
GLM-5.2 (max) | 262k | 开放 | 53 | $0.67 | 190 | 0.73 | 13.88 | 10.52 | ||
GLM-5.2 (max) | 262k | 开放 | 53 | $1.57 | 68 | 1.84 | 38.88 | 29.63 | ||
GLM-5.2 (max) | 1M | 开放 | 53 | -- | 87 | 0.87 | 29.76 | 23.11 | ||
GLM-5.2 (max) (NVFP4) | 1M | 开放 | 53 | $0.36 | 180 | 0.99 | 14.85 | 11.08 | ||
GLM-5.2 (max) | 1M | 开放 | 53 | $0.48 | 92 | 1.26 | 28.51 | 21.80 | ||
GLM-5.2 (max) (FP8) | 1.05M | 开放 | 53 | -- | -- | -- | -- | -- | ||
GLM-5.2 (max) (FP8) | 1.05M | 开放 | 53 | $0.62 | 86 | 1.94 | 31.14 | 23.36 | ||
Claude Opus 5 (low) | 1M | 专有 | 52 | $0.43 | 53 | 3.47 | 12.94 | -- | ||
Claude Opus 5 (low) | 1M | 专有 | 52 | $0.43 | 54 | 6.06 | 15.40 | -- | ||
Claude Opus 5 (low) | 1M | 专有 | 52 | $0.55 | 55 | 3.68 | 12.78 | -- | ||
GPT-5.6 Luna (max) | 1M | 专有 | 52 | $0.05 | 152 | 157.72 | 161.00 | -- | ||
GPT-5.6 Luna (max) | 1M | 专有 | 52 | $0.34 | 222 | 91.61 | 93.86 | -- | ||
DeepSeek V4 Flash 0731 (max) | 262k | 开放 | 52 | $0.12 | 123 | 1.70 | 21.95 | 16.20 | ||
DeepSeek V4 Flash 0731 (max) (FAST) | 1M | 开放 | 52 | $0.17 | 165 | 6.05 | 21.21 | 12.13 | ||
DeepSeek V4 Flash 0731 (max) | 1.05M | 开放 | 52 | $0.07 | 69 | 1.10 | 37.42 | 29.06 | ||
DeepSeek V4 Flash 0731 (max) | 1.05M | 开放 | 52 | $0.20 | 250 | 1.77 | 11.76 | 7.99 | ||
DeepSeek V4 Flash 0731 (max) | 1M | 开放 | 52 | $0.03 | 121 | 1.41 | 22.13 | 16.58 | ||
DeepSeek V4 Flash 0731 (max) | 1M | 开放 | 52 | $0.06 | 53 | 0.94 | 47.90 | 37.57 | ||
DeepSeek V4 Flash 0731 (max) | 1.05M | 开放 | 52 | $0.13 | 67 | 1.69 | 39.16 | 29.97 | ||
DeepSeek V4 Flash 0731 (max) | 1.05M | 开放 | 52 | $0.08 | 112 | 0.96 | 23.32 | 17.89 | ||
DeepSeek V4 Flash 0731 (max) | 1.05M | 开放 | 52 | $0.09 | 107 | 1.63 | 25.02 | 18.71 | ||
DeepSeek V4 Flash 0731 (max) | 1.05M | 开放 | 52 | $0.14 | 291 | 0.74 | 9.33 | 6.87 | ||
DeepSeek V4 Flash 0731 (max) | 1.05M | 开放 | 52 | $0.15 | 124 | 2.87 | 23.03 | 16.13 | ||
DeepSeek V4 Flash 0731 (max) | 1.05M | 开放 | 52 | $0.08 | 125 | 1.36 | 21.29 | 15.95 | ||
DeepSeek V4 Flash 0731 (max) | 1M | 开放 | 52 | $0.20 | 285 | 0.67 | 9.46 | 7.03 | ||
Gemini 3.6 Flash AI Studio | 1M | 专有 | 52 | $0.56 | 225 | 19.69 | 21.91 | -- | ||
Gemini 3.7 Flash (low) AI Studio | 1M | 专有 | 51 | $0.16 | 253 | 0.74 | 2.71 | -- | ||
GPT-5.6 Sol (low) | 1M | 专有 | 51 | $0.23 | 59 | 3.74 | 12.20 | -- | ||
GPT-5.6 Terra (high) | 1M | 专有 | 50 | $0.22 | 96 | 4.02 | 9.23 | -- | ||
GPT-5.6 Luna (xhigh) | 1M | 专有 | 50 | $0.03 | 148 | 50.52 | 53.90 | -- | ||
Kimi K3 (low) | 1.05M | 开放 | 48 | $0.24 | 37 | 3.32 | 71.66 | 54.67 | ||
Gemini 3.1 Pro Preview (AI Studio) | 1M | 专有 | 48 | $0.33 | 113 | 28.33 | 32.74 | -- | ||
Gemini 3.1 Pro Preview (Vertex) | 1M | 专有 | 48 | $0.43 | 127 | 23.65 | 27.59 | -- | ||
GPT-5.6 Luna (high) | 1M | 专有 | 47 | $0.02 | 153 | 16.98 | 20.24 | -- | ||
GPT-5.6 Terra (medium) | 1M | 专有 | 47 | $0.12 | 89 | 1.76 | 7.41 | -- | ||
Gemini 3.5 Flash (medium) | 1M | 专有 | 47* | -- | 187 | 21.10 | 23.77 | -- | ||
GPT-5.3 Codex (xhigh) | 400k | 专有 | 46* | -- | 111 | 78.88 | 83.38 | -- | ||
MiniMax-M3 (MXFP8) | 1M | 开放 | 45 | $0.18 | 97 | 1.28 | 26.96 | 20.54 | ||
MiniMax-M3 | 262k | 开放 | 45 | $0.11 | 199 | 1.15 | 13.74 | 10.07 | ||
MiniMax-M3 | 1M | 开放 | 45 | $0.21 | 51 | 0.93 | 49.71 | 39.02 | ||
MiniMax-M3 | 1M | 开放 | 45 | $0.13 | 83 | 1.55 | 31.52 | 23.98 | ||
MiniMax-M3 | 1M | 开放 | 45 | $0.18 | 232 | 0.52 | 11.32 | 8.64 | ||
MiniMax-M3 | 524k | 开放 | 45 | $0.13 | 28 | 0.99 | 89.18 | 70.55 | ||
MiniMax-M3 (FP8) | 1.05M | 开放 | 45 | $0.45 | 213 | 1.90 | 13.65 | 9.40 | ||
MiniMax-M3 (FP8) | 1M | 开放 | 45 | $0.12 | 80 | 1.67 | 32.82 | 24.93 | ||
MiniMax-M3 | 512k | 开放 | 45 | $0.14 | 86 | 1.65 | 30.84 | 23.36 | ||
MiniMax-M3 | 1.05M | 开放 | 45 | -- | -- | -- | -- | -- | ||
DeepSeek V4 Pro (max) (FP4) | 1.05M | 开放 | 45 | $0.27 | 33 | 1.44 | 147.51 | 131.09 | ||
DeepSeek V4 Pro (max) | 1M | 开放 | 45 | $0.89 | 91 | 1.50 | 55.25 | 48.23 | ||
DeepSeek V4 Pro (max) | 1.05M | 开放 | 45 | $0.23 | 72 | 1.77 | 69.83 | 61.08 | ||
DeepSeek V4 Pro (max) (FP8) | 1.05M | 开放 | 45 | $0.31 | 57 | 1.65 | 86.85 | 76.46 | ||
DeepSeek V4 Pro (max) | 1.05M | 开放 | 45 | -- | -- | -- | -- | -- | ||
DeepSeek V4 Pro (max) | 1M | 开放 | 45 | $0.90 | 90 | 1.35 | 55.55 | 48.64 | ||
DeepSeek V4 Pro (max) | 1.05M | 开放 | 45 | $0.38 | 69 | 1.62 | 72.27 | 63.40 | ||
DeepSeek V4 Pro (max) | 131k | 开放 | 45 | -- | -- | -- | -- | -- | ||
DeepSeek V4 Pro (max) | 1M | 开放 | 45 | $0.05 | 71 | 1.65 | 70.74 | 62.00 | ||
DeepSeek V4 Pro (high) (FP4) | 65.5k | 开放 | 44 | $0.26 | 33 | 1.69 | 76.55 | 59.84 | ||
DeepSeek V4 Pro (high) | 1.05M | 开放 | 44 | $0.36 | 68 | 2.42 | 39.06 | 29.29 | ||
DeepSeek V4 Pro (high) | 1M | 开放 | 44 | $0.94 | 80 | 1.56 | 32.86 | 25.03 | ||
DeepSeek V4 Pro (high) | 1M | 开放 | 44 | $0.04 | 70 | 1.75 | 37.31 | 28.42 | ||
DeepSeek V4 Pro (high) (FP8) | 1.05M | 开放 | 44 | $0.29 | 60 | 2.08 | 43.43 | 33.05 | ||
DeepSeek V4 Pro (high) | 1M | 开放 | 44 | $0.94 | 84 | 1.34 | 30.93 | 23.65 | ||
DeepSeek V4 Pro (high) | 1.05M | 开放 | 44 | $0.20 | 73 | 1.84 | 35.83 | 27.17 | ||
DeepSeek V4 Pro (high) | 131k | 开放 | 44 | $0.68 | 158 | 0.61 | 16.40 | 12.62 | ||
Kimi K2.7 Code (FP4) | 256k | 开放 | 43 | $0.64 | 52 | 1.82 | 54.40 | 42.95 | ||
Kimi K2.7 Code (FP8) | 65.5k | 开放 | 43 | $0.46 | 39 | 5.87 | 76.58 | 57.76 | ||
Kimi K2.7 Code | 262k | 开放 | 43 | $0.28 | 310 | 1.13 | 9.92 | 7.18 | ||
Kimi K2.7 Code | 262k | 开放 | 43 | $0.18 | 77 | 1.27 | 36.51 | 28.78 | ||
Kimi K2.7 Code | 262k | 开放 | 43 | $0.30 | 272 | 0.45 | 10.48 | 8.20 | ||
Kimi K2.7 Code | 262k | 开放 | 43 | $0.29 | -- | -- | -- | -- | ||
Kimi K2.7 Code | 262k | 开放 | 43 | $0.23 | 38 | 3.73 | 74.77 | 58.03 | ||
Kimi K2.7 Code | 262k | 开放 | 43 | $0.18 | 113 | 0.67 | 24.84 | 19.73 | ||
Kimi K2.7 Code | 256k | 开放 | 43 | $0.22 | 37 | 2.92 | 77.59 | 60.99 | ||
MiMo-V2.5-Pro | 1.05M | 开放 | 43 | $0.04 | 50 | 2.55 | 52.95 | 40.32 | ||
MiMo-V2.5-Pro | 1.05M | 开放 | 43 | -- | -- | -- | -- | -- | ||
MiMo-V2.5-Pro | 65.5k | 开放 | 43 | $0.18 | 128 | 0.81 | 20.27 | 15.57 | ||
MiMo-V2.5-Pro | 1M | 开放 | 43 | $0.03 | 46 | 3.57 | 57.61 | 43.23 | ||
Claude Sonnet 5 (Non-reasoning) | 1M | 专有 | 43 | $0.42 | 67 | 1.82 | 9.30 | -- | ||
Claude Sonnet 5 (Non-reasoning) | 1M | 专有 | 43 | $0.61 | 66 | 2.91 | 10.46 | -- | ||
Inkling (FP8) | 131k | 开放 | 42 | $0.43 | 103 | 0.59 | 24.85 | 19.41 | ||
Inkling | 262k | 开放 | 42 | $0.34 | 233 | 0.63 | 11.35 | 8.58 | ||
Inkling | 524k | 开放 | 42 | $0.62 | 89 | 0.66 | 28.84 | 22.54 | ||
Inkling | 262k | 开放 | 42 | $0.34 | 73 | 1.84 | 35.92 | 27.27 | ||
Hy3 (FP8) | 256k | 开放 | 42 | $0.03 | 68 | 2.85 | 39.74 | 29.51 | ||
Hy3 | 262k | 开放 | 42 | $0.04 | 68 | 2.90 | 39.63 | 29.38 | ||
Hy3 (FP8) | 262k | 开放 | 42 | $0.04 | 53 | 0.74 | 47.51 | 37.42 | ||
Hy3 | 262k | 开放 | 42 | $0.03 | 68 | 3.41 | 40.10 | 29.35 | ||
Nex-N2-Pro (FP8) | 262k | 开放 | 42* | -- | 134 | 1.73 | 20.39 | 14.93 | ||
GPT-5.6 Sol (Non-reasoning) | 1M | 专有 | 42 | $0.24 | 59 | 1.38 | 9.87 | -- | ||
GPT-5.6 Sol (Non-reasoning) | 1M | 专有 | 42 | $0.30 | 104 | 2.22 | 7.03 | -- | ||
Solar Pro 4 | 512k | 专有 | 42 | $0.22 | 57 | 2.33 | 46.53 | 35.36 | ||
GPT-5.6 Terra (low) | 1M | 专有 | 41 | $0.09 | 90 | 1.79 | 7.37 | -- | ||
Inkling Small | 256k | 开放 | 41 | $0.07 | 114 | 1.79 | 23.78 | 17.59 | ||
Inkling Small | 262k | 开放 | 41 | $0.29 | 345 | 0.52 | 7.78 | 5.80 | ||
Inkling Small | 524k | 开放 | 41 | $0.15 | 87 | 0.81 | 29.58 | 23.02 | ||
Agnes 2.5 Pro Alpha | 1M | 专有 | 40 | -- | 131 | 2.18 | 21.20 | 15.21 | ||
Qwen3.7 Plus | 1M | 专有 | 39 | $0.24 | 56 | 2.21 | 46.68 | 35.58 | ||
GPT-5.6 Luna (medium) | 1M | 专有 | 39 | $0.01 | 141 | 2.33 | 5.88 | -- | ||
Nemotron 3 Ultra | 262k | 开放 | 38 | $0.16 | 211 | 1.08 | 14.26 | 10.80 | ||
Nemotron 3 Ultra | 262k | 开放 | 38 | -- | -- | -- | -- | -- | ||
Nemotron 3 Ultra | 512k | 开放 | 38 | $0.40 | 151 | 0.66 | 19.00 | 15.04 | ||
Nemotron 3 Ultra | 262k | 开放 | 38 | $0.23 | 115 | 3.69 | 27.84 | 19.80 | ||
Nemotron 3 Ultra BF16 | 262k | 开放 | 38 | -- | 83 | 5.43 | 38.87 | 27.42 | ||
Nemotron 3 Ultra | 256k | 开放 | 38 | $0.60 | 351 | 2.60 | 10.50 | 6.47 | ||
Nemotron 3 Ultra | 262k | 开放 | 38 | $0.25 | -- | -- | -- | -- | ||
Nemotron 3 Ultra | 262k | 开放 | 38 | $0.22 | -- | -- | -- | -- | ||
MiMo-V2.5 | 1.05M | 开放 | 38 | $0.01 | 96 | 1.71 | 27.82 | 20.88 | ||
MiMo-V2.5 | 1M | 开放 | 38 | $0.04 | 65 | 1.79 | 40.25 | 30.77 | ||
MiMo-V2.5 | 262k | 开放 | 38 | $0.08 | 30 | 1.25 | 85.38 | 67.30 | ||
MiMo-V2.5 | 1.05M | 开放 | 38 | $0.02 | -- | -- | -- | -- | ||
MiMo-V2.5 | 1M | 开放 | 38 | $0.01 | 101 | 1.94 | 26.69 | 19.80 | ||
Ling 3.0 Flash | 131k | 开放 | 38 | $0.04 | 65 | 0.95 | 39.69 | 30.99 | ||
Ling 3.0 Flash | 262k | 开放 | 38 | $0.04 | 374 | 2.06 | 8.74 | 5.35 | ||
Qwen3.6 27B FP8 | 262k | 开放 | 38 | $0.19 | 52 | 1.01 | 119.56 | 108.96 | ||
Qwen3.6 27B | 131k | 开放 | 38 | $0.27 | 448 | 1.24 | 15.03 | 12.68 | ||
Qwen3.6 27B | 262k | 开放 | 38 | $0.29 | 54 | 2.91 | 116.44 | 104.34 | ||
Qwen3.6 27B | 262k | 开放 | 38 | $0.29 | 55 | 3.80 | 116.88 | 103.92 | ||
Qwen3.6 27B | 262k | 开放 | 38 | $0.07 | 183 | 1.37 | 35.20 | 31.09 | ||
Qwen3.6 27B (FP8) | 262k | 开放 | 38 | $0.19 | 40 | 3.59 | 158.74 | 142.58 | ||
Gemini 3.5 Flash-Lite AI Studio | 1M | 专有 | 37 | $0.10 | 368 | 10.11 | 11.47 | -- | ||
Grok 4.3 (medium) | 524k | 专有 | 37* | -- | 197 | 9.51 | 12.05 | -- | ||
Grok 4.3 (medium) | 200k | 专有 | 37* | -- | 157 | 12.95 | 16.14 | -- | ||
Grok 4.3 (medium) | 1M | 专有 | 37* | -- | 117 | 15.29 | 19.56 | -- | ||
Grok 4.3 (low) | 524k | 专有 | 36* | -- | 191 | 4.78 | 7.40 | -- | ||
Grok 4.3 (low) | 200k | 专有 | 36* | -- | 167 | 5.36 | 8.36 | -- | ||
Grok 4.3 (low) | 1M | 专有 | 36* | -- | 117 | 7.54 | 11.82 | -- | ||
Gemini 3.5 Flash (minimal) AI Studio | 1M | 专有 | 36* | -- | 161 | 0.89 | 3.99 | -- | ||
Claude Sonnet 4.6 (Non-reasoning, Low Effort) | 1M | 专有 | 35* | -- | 44 | 1.92 | 13.33 | -- | ||
Muse Glimmer (high) | 131k | 开放 | 35 | $0.08 | 81 | 0.80 | 31.61 | 24.65 | ||
Muse Glimmer (high) | 131k | 开放 | 35 | $0.07 | 139 | 0.90 | 18.92 | 14.42 | ||
GLM-5.2 (FAST) | 524k | 开放 | 35 | -- | 185 | 1.44 | 4.14 | -- | ||
GLM-5.2 (FP4) | 432k | 开放 | 35 | -- | 177 | 1.15 | 3.97 | -- | ||
GLM-5.2 | 262k | 开放 | 35 | -- | 69 | 1.76 | 9.00 | -- | ||
GLM-5.2 | 205k | 开放 | 35 | -- | 175 | 6.28 | 9.13 | -- | ||
GLM-5.2 (FP8) | 1.05M | 开放 | 35 | -- | 86 | 2.11 | 7.94 | -- | ||
GPT-5.6 Terra (Non-reasoning) | 1M | 专有 | 35 | $0.10 | 96 | 0.81 | 6.01 | -- | ||
GPT-5.6 Terra (Non-reasoning) | 1M | 专有 | 35 | $0.20 | 141 | 0.62 | 4.17 | -- | ||
KAT-Coder-Pro V2 | 256k | 专有 | 35* | -- | 104 | 1.51 | 6.33 | -- | ||
KAT-Coder-Pro V2 | 256k | 专有 | 35* | -- | -- | -- | -- | -- | ||
LongCat 2.0 (FP8) | 1M | 开放 | 34* | -- | 40 | 2.92 | 65.16 | 49.79 | ||
Qwen3.5 397B A17B | 256k | 开放 | 34 | $0.36 | -- | -- | -- | -- | ||
Qwen3.5 397B A17B | 262k | 开放 | 34 | $0.36 | 82 | 2.19 | 46.98 | 38.71 | ||
Qwen3.5 397B A17B | 262k | 开放 | 34 | $0.12 | 88 | 2.18 | 44.11 | 36.24 | ||
Qwen3.5 397B A17B | 262k | 开放 | 34 | $0.23 | 68 | 1.22 | 55.66 | 47.05 | ||
Qwen3.5 397B A17B (FP8) | 262k | 开放 | 34 | $0.23 | 92 | 2.31 | 42.44 | 34.68 | ||
Qwen3.5 397B A17B (FP8) | 262k | 开放 | 34 | $0.19 | 39 | 1.14 | 94.72 | 80.88 | ||
Qwen3.5 397B A17B | 262k | 开放 | 34 | $0.12 | 10 | 1.73 | 379.12 | 326.20 | ||
Qwen3.5 397B A17B (FP8) | 262k | 开放 | 34 | $0.36 | -- | -- | -- | -- | ||
Qwen3.5 397B A17B (Base, FP4) | 262k | 开放 | 34 | $0.36 | 131 | 2.05 | 30.18 | 24.32 | ||
Qwen3.5 397B A17B | 262k | 开放 | 34 | $0.36 | 83 | 1.45 | 45.88 | 38.40 | ||
GPT-5.6 Luna (low) | 1M | 专有 | 34 | $0.01 | 142 | 1.85 | 5.37 | -- | ||
Qwen3.5 122B A10B (FP8) | 262k | 开放 | 33 | $0.16 | 67 | 1.92 | 39.16 | 29.79 | ||
Qwen3.5 122B A10B (FP4) | 262k | 开放 | 33 | $0.12 | 138 | 0.77 | 18.84 | 14.45 | ||
Qwen3.5 122B A10B | 262k | 开放 | 33 | $0.25 | 136 | 2.36 | 20.77 | 14.73 | ||
Qwen3.5 122B A10B (FP8) | 262k | 开放 | 33 | $0.25 | -- | -- | -- | -- | ||
Qwen3.5 122B A10B | 262k | 开放 | 33 | $0.25 | 108 | 2.04 | 25.19 | 18.52 | ||
Qwen3.5 397B A17B | 262k | 开放 | 33* | -- | 89 | 2.16 | 7.78 | -- | ||
Qwen3.5 397B A17B (FP8) | 262k | 开放 | 33* | -- | 42 | 0.99 | 12.84 | -- | ||
Qwen3.5 397B A17B | 262k | 开放 | 33* | -- | 82 | 2.07 | 8.16 | -- | ||
Qwen3.5 397B A17B (Base, FP4) | 262k | 开放 | 33* | -- | 130 | 2.07 | 5.91 | -- | ||
Qwen3.5 397B A17B | 262k | 开放 | 33* | -- | 7 | 1.79 | 68.78 | -- | ||
Qwen3.5 397B A17B | 262k | 开放 | 33* | -- | 86 | 1.40 | 7.24 | -- | ||
Qwen3.6 35B A3B | 229k | 开放 | 32 | $0.29 | 146 | 2.19 | 42.49 | 36.88 | ||
Qwen3.6 35B A3B (FP8) | 262k | 开放 | 32 | $0.17 | 57 | 2.24 | 104.84 | 93.90 | ||
Qwen3.6 35B A3B | 262k | 开放 | 32 | $0.23 | 137 | 1.00 | 43.98 | 39.34 | ||
Qwen3.6 35B A3B (FP8) | 262k | 开放 | 32 | $0.09 | 26 | 0.81 | 231.46 | 211.08 | ||
Qwen3.6 35B A3B (FP8) | 262k | 开放 | 32 | $0.47 | -- | -- | -- | -- | ||
Qwen3.6 35B A3B FP8 | 262k | 开放 | 32 | $0.19 | -- | -- | -- | -- | ||
Qwen3.6 35B A3B | 262k | 开放 | 32 | $0.08 | 74 | 1.30 | 80.94 | 72.89 | ||
Qwen3.6 35B A3B | 205k | 开放 | 32 | $0.19 | 140 | 1.48 | 43.68 | 38.62 | ||
DeepSeek V4 Pro | 1M | 开放 | 32* | -- | 94 | 1.66 | 6.97 | -- | ||
DeepSeek V4 Pro | 1.05M | 开放 | 32* | -- | 84 | 1.34 | 7.26 | -- | ||
DeepSeek V4 Pro | 1M | 开放 | 32* | -- | 73 | 1.73 | 8.60 | -- | ||
G9v3-39A5B | 131k | 开放 | 32* | -- | -- | -- | -- | -- | ||
Qwen3.5 Omni Plus | 256k | 专有 | 31* | -- | 50 | 2.50 | 12.59 | -- | ||
Qwen3.6 27B | 131k | 开放 | 31 | $0.39 | 481 | 1.07 | 2.11 | -- | ||
Qwen3.6 27B | 262k | 开放 | 31 | $0.10 | 191 | 1.29 | 3.91 | -- | ||
Qwen3.6 27B FP8 | 262k | 开放 | 31 | $0.23 | 42 | 1.14 | 12.96 | -- | ||
Qwen3.6 27B | 262k | 开放 | 31 | $0.40 | 57 | 3.94 | 12.78 | -- | ||
Qwen3.6 27B | 262k | 开放 | 31 | $0.40 | 59 | 2.93 | 11.41 | -- | ||
Ring-2.6-1T | 262k | 开放 | 31* | -- | 121 | 3.27 | 23.95 | 16.55 | ||
o3 | 200k | 专有 | 31* | -- | 113 | 37.52 | 41.96 | -- | ||
o3 | 200k | 专有 | 31* | -- | -- | -- | -- | -- | ||
o3 | 200k | 专有 | 31* | -- | 118 | 7.37 | 11.61 | -- | ||
Step 3.7 Flash | 256k | 开放 | 31 | $0.09 | 399 | 0.84 | 7.10 | 5.01 | ||
Step 3.7 Flash | 256k | 开放 | 31 | -- | 176 | 0.68 | 14.87 | 11.35 | ||
Mistral Medium 3.5 | 262k | 开放 | 30 | $0.46 | 69 | 2.35 | 38.39 | 28.83 | ||
Claude 4.5 Haiku | 200k | 专有 | 30 | $0.24 | 110 | 11.73 | 16.29 | -- | ||
Claude 4.5 Haiku Vertex | 200k | 专有 | 30 | $0.25 | 91 | 23.34 | 28.84 | -- | ||
Claude 4.5 Haiku | 200k | 专有 | 30 | $0.22 | 92 | 15.77 | 21.20 | -- | ||
Claude 4.5 Haiku | 200k | 专有 | 30 | $0.54 | 89 | 16.39 | 22.02 | -- | ||
Gemma 4 31B | 262k | 开放 | 30 | -- | 35 | 1.95 | 65.09 | 49.02 | ||
Gemma 4 31B (FP8) | 262k | 开放 | 30 | $0.04 | -- | -- | -- | -- | ||
Gemma 4 31B | 262k | 开放 | 30 | $0.04 | -- | -- | -- | -- | ||
Gemma 4 31B | 131k | 开放 | 30 | $0.24 | 1,510 | 1.11 | 2.59 | 1.15 | ||
Gemma 4 31B (AI Studio) | 262k | 开放 | 30 | $0.00 | 36 | 1.12 | 63.95 | 48.78 | ||
Gemma 4 31B | 262k | 开放 | 30 | $0.02 | 14 | 3.59 | 163.18 | 123.90 | ||
Gemma 4 31B | 262k | 开放 | 30 | $0.01 | 30 | 1.54 | 77.14 | 58.69 | ||
Gemma 4 31B (FP8) | 262k | 开放 | 30 | $0.03 | 49 | 3.82 | 49.16 | 35.20 | ||
Gemma 4 31B | 131k | 开放 | 30 | $0.10 | 199 | 2.68 | 13.92 | 8.72 | ||
Gemma 4 31B | 256k | 开放 | 30 | $0.04 | 131 | 2.98 | 20.06 | 13.26 | ||
Gemma 4 31B | 131k | 开放 | 30 | $0.04 | 135 | 0.94 | 17.44 | 12.81 | ||
Gemma 4 31B (FP8) | 262k | 开放 | 30 | $0.10 | 46 | 1.30 | 50.20 | 37.97 | ||
DeepSeek V4 Flash | 1.05M | 开放 | 29* | -- | -- | -- | -- | -- | ||
DeepSeek V4 Flash | 1M | 开放 | 29* | -- | 311 | 0.65 | 2.25 | -- | ||
DeepSeek V4 Flash | 1M | 开放 | 29* | -- | 68 | 1.62 | 8.98 | -- | ||
GPT-5.5 Instant (June 2026) | 400k | 专有 | 29 | $0.54 | 148 | 1.09 | 17.97 | 13.51 | ||
MiMo-V2.5-Pro | 1.05M | 开放 | 28* | -- | -- | -- | -- | -- | ||
MiMo-V2.5-Pro | 65.5k | 开放 | 28* | -- | 129 | 0.74 | 4.62 | -- | ||
MiMo-V2.5-Pro | 131k | 开放 | 28* | -- | 47 | 3.23 | 13.91 | -- | ||
MiMo-V2.5-Pro | 1.05M | 开放 | 28* | -- | 47 | 2.46 | 13.07 | -- | ||
Qwen3.5 122B A10B (FP4) | 262k | 开放 | 28 | $0.10 | 151 | 0.74 | 4.05 | -- | ||
Qwen3.5 122B A10B | 262k | 开放 | 28 | $0.20 | 146 | 2.37 | 5.79 | -- | ||
GPT-5.6 Luna (Non-reasoning) | 1M | 专有 | 27 | $0.01 | 160 | 0.79 | 3.91 | -- | ||
GPT-5.6 Luna (Non-reasoning) | 1M | 专有 | 27 | $0.09 | 224 | 0.69 | 2.93 | -- | ||
Gemma 4 26B A4B | 1M | 开放 | 26 | $0.02 | 160 | 1.02 | 16.60 | 12.47 | ||
Gemma 4 26B A4B | 262k | 开放 | 26 | $0.04 | 34 | 4.23 | 76.75 | 58.01 | ||
Gemma 4 26B A4B AI Studio | 262k | 开放 | 26 | $0.00 | 47 | 1.07 | 53.98 | 42.32 | ||
Gemma 4 26B A4B | 256k | 开放 | 26 | $0.03 | 108 | 1.38 | 24.45 | 18.45 | ||
Gemma 4 26B A4B | 262k | 开放 | 26 | $0.03 | 21 | 1.05 | 118.03 | 93.59 | ||
Gemma 4 26B A4B | 256k | 开放 | 26 | $0.03 | 49 | 1.63 | 52.72 | 40.87 | ||
Gemma 4 26B A4B | 256k | 开放 | 26 | $0.20 | -- | -- | -- | -- | ||
Gemma 4 26B A4B (FP8) | 1.05M | 开放 | 26 | $0.04 | -- | -- | -- | -- | ||
Nemotron 3 Super | 262k | 开放 | 26 | $0.23 | 142 | 0.97 | 18.52 | 14.04 | ||
Nemotron 3 Super | 256k | 开放 | 26 | $0.33 | 356 | 1.87 | 8.90 | 5.62 | ||
Nemotron 3 Super | 262k | 开放 | 26 | $0.10 | 59 | 19.42 | 61.52 | 33.69 | ||
Grok 4.3 (Non-reasoning) | 524k | 专有 | 25 | $0.91 | 165 | 1.08 | 4.11 | -- | ||
Grok 4.3 (Non-reasoning) | 1M | 专有 | 25 | $0.29 | 97 | 0.80 | 5.94 | -- | ||
Qwen3.6 35B A3B (FP8) | 262k | 开放 | 25* | -- | -- | -- | -- | -- | ||
Qwen3.6 35B A3B | 262k | 开放 | 25* | -- | 144 | 0.90 | 4.37 | -- | ||
Qwen3.6 35B A3B (FP8) | 262k | 开放 | 25* | -- | 22 | 0.80 | 23.13 | -- | ||
Qwen3.6 35B A3B | 205k | 开放 | 25* | -- | 166 | 1.58 | 4.59 | -- | ||
Qwen3.6 35B A3B FP8 | 262k | 开放 | 25* | -- | -- | -- | -- | -- | ||
Qwen3.6 35B A3B (FP8) | 262k | 开放 | 25* | -- | 76 | 1.04 | 7.65 | -- | ||
Qwen3.6 35B A3B | 262k | 开放 | 25* | -- | 154 | 2.33 | 5.57 | -- | ||
Ling 3.0 Tiny | 262k | 开放 | 25 | $0.00 | 369 | 1.27 | 8.05 | 5.42 | ||
Ling 3.0 Tiny | 262k | 开放 | 25 | $0.00 | 159 | 2.71 | 18.44 | 12.59 | ||
Qwen3.5 35B A3B FP8 | 262k | 开放 | 24 | $0.07 | 113 | 0.55 | 4.98 | -- | ||
Qwen3.5 35B A3B | 262k | 开放 | 24 | $0.24 | 171 | 2.11 | 5.03 | -- | ||
Claude 4.5 Haiku | 200k | 专有 | 24* | -- | 89 | 1.25 | 6.89 | -- | ||
Claude 4.5 Haiku | 200k | 专有 | 24* | -- | 107 | 1.06 | 5.75 | -- | ||
Claude 4.5 Haiku | 200k | 专有 | 24* | -- | 88 | 0.82 | 6.48 | -- | ||
Claude 4.5 Haiku Vertex | 200k | 专有 | 24* | -- | 90 | 0.70 | 6.24 | -- | ||
gpt-oss-120b (high) | 131k | 开放 | 24 | $0.06 | 476 | 0.71 | 5.96 | 4.20 | ||
gpt-oss-120b (high) | 116k | 开放 | 24 | $0.15 | 80 | 0.78 | 32.11 | 25.07 | ||
gpt-oss-120b (high) | 128k | 开放 | 24 | $0.08 | 163 | 1.29 | 16.65 | 12.29 | ||
gpt-oss-120b (high) | 131k | 开放 | 24 | $0.02 | 21 | 1.53 | 119.27 | 94.19 | ||
gpt-oss-120b (high) | 131k | 开放 | 24 | $0.02 | 31 | 1.47 | 81.04 | 63.66 | ||
gpt-oss-120b (high) | 131k | 开放 | 24 | $0.07 | 310 | 0.83 | 8.90 | 6.46 | ||
gpt-oss-120b (high) | 131k | 开放 | 24 | $0.03 | 104 | 0.88 | 24.85 | 19.18 | ||
gpt-oss-120b (high) (Turbo) | 131k | 开放 | 24 | $0.07 | 165 | 0.70 | 15.90 | 12.16 | ||
gpt-oss-120b (high) | 131k | 开放 | 24 | $0.15 | 1,742 | 0.52 | 1.96 | 1.15 | ||
gpt-oss-120b (high) Vertex | 131k | 开放 | 24 | $0.04 | 166 | 0.75 | 15.85 | 12.08 | ||
gpt-oss-120b (high) | 131k | 开放 | 24 | $0.04 | -- | -- | -- | -- | ||
gpt-oss-120b (high) | 128k | 开放 | 24 | $0.05 | 207 | 0.23 | 12.32 | 9.67 | ||
gpt-oss-120b (high) | 131k | 开放 | 24 | $0.10 | 701 | 1.04 | 4.61 | 2.85 | ||
gpt-oss-120b (high) | 131k | 开放 | 24 | -- | 117 | 0.95 | 22.23 | 17.03 | ||
gpt-oss-120b (high) (Base) | 128k | 开放 | 24 | $0.07 | 296 | 0.98 | 9.42 | 6.76 | ||
gpt-oss-120b (high) | 131k | 开放 | 24 | $0.11 | -- | -- | -- | -- | ||
gpt-oss-120b (high) | 131k | 开放 | 24 | $0.07 | 69 | 1.03 | 37.48 | 29.16 | ||
gpt-oss-120b (high) | 131k | 开放 | 24 | $0.07 | 70 | 0.64 | 36.45 | 28.65 | ||
Nemotron 3.5 Lightning (NVFP4) | 28.7k | 开放 | 24 | $0.06 | 472 | 0.47 | 5.77 | 4.23 | ||
Nemotron 3.5 Lightning (BF16) | 1.05M | 开放 | 24 | $0.07 | 290 | 1.78 | 10.41 | 6.90 | ||
Nemotron 3.5 Lightning (BF16) | 262k | 开放 | 24 | $0.08 | 165 | 0.98 | 16.15 | 12.14 | ||
Nemotron 3.5 Lightning (BF16) | 65.5k | 开放 | 24 | $0.00 | 292 | 1.79 | 10.35 | 6.85 | ||
Nemotron 3.5 Lightning (BF16) | 262k | 开放 | 24 | $0.03 | 499 | 0.58 | 5.59 | 4.01 | ||
Command A+ | 200k | 开放 | 23 | $0.00 | 199 | 0.40 | 12.98 | 10.07 | ||
Gemma 4 31B | 262k | 开放 | 22 | $0.03 | 24 | 3.15 | 23.64 | -- | ||
Gemma 4 31B | 131k | 开放 | 22 | $0.28 | 1,423 | 1.39 | 1.74 | -- | ||
Gemma 4 31B | 262k | 开放 | 22 | $0.04 | 123 | 3.51 | 7.58 | -- | ||
Gemma 4 31B (FP8) | 262k | 开放 | 22 | $0.11 | 53 | 1.27 | 10.77 | -- | ||
Gemma 4 31B (FP8) | 262k | 开放 | 22 | $0.04 | 61 | 3.50 | 11.73 | -- | ||
Gemma 4 31B (FP8) | 262k | 开放 | 22 | $0.04 | 23 | 1.03 | 22.46 | -- | ||
Gemma 4 31B | 131k | 开放 | 22 | $0.11 | 202 | 2.38 | 4.86 | -- | ||
Gemma 4 31B | 262k | 开放 | 22 | $0.04 | 17 | 1.70 | 31.04 | -- | ||
Gemma 4 12B | 262k | 开放 | 22* | -- | 110 | 2.36 | 25.08 | 18.18 | ||
Nova 2.0 Pro Preview (medium) | 256k | 专有 | 22* | -- | 115 | 16.60 | 38.41 | 17.45 | ||
Mercury 2 | 128k | 专有 | 22 | $0.08 | 820 | 3.34 | 3.95 | -- | ||
Qwen3.5 9B (FP8) | 262k | 开放 | 22 | $0.18 | 77 | 2.47 | 35.09 | 26.10 | ||
Qwen3.5 9B (FP8) | 262k | 开放 | 22 | $0.30 | 95 | 0.70 | 26.88 | 20.94 | ||
Qwen3 Coder Next | 128k | 开放 | 21 | $0.47 | 111 | 1.57 | 6.08 | -- | ||
Qwen3 Coder Next (FP8) | 262k | 开放 | 21 | $0.08 | 110 | 0.90 | 5.45 | -- | ||
Qwen3 Coder Next (FP8) | 262k | 开放 | 21 | $0.47 | -- | -- | -- | -- | ||
Qwen3 Coder Next (FP8) | 262k | 开放 | 21 | $0.21 | 156 | 1.82 | 5.03 | -- | ||
Nova 2.0 Omni (medium) | 1M | 专有 | 21* | -- | -- | -- | -- | -- | ||
Nova 2.0 Omni (medium) | 1M | 专有 | 21* | -- | -- | -- | -- | -- | ||
Apriel-v1.6-15B-Thinker | 131k | 开放 | 21* | -- | -- | -- | -- | -- | ||
Nova 2.0 Lite (high) | 1M | 专有 | 21* | -- | -- | -- | -- | -- | ||
Nova 2.0 Lite (high) | 1M | 专有 | 21* | -- | 157 | 15.31 | 31.26 | 12.76 | ||
Qwen3.5 9B (FP8) | 262k | 开放 | 21* | -- | 92 | 0.85 | 6.29 | -- | ||
Gemma 4 26B A4B | 1M | 开放 | 20* | -- | 177 | 1.04 | 3.87 | -- | ||
Gemma 4 26B A4B | 262k | 开放 | 20* | -- | 39 | 1.60 | 14.36 | -- | ||
Gemma 4 26B A4B (FP8) | 1.05M | 开放 | 20* | -- | -- | -- | -- | -- | ||
Gemma 4 26B A4B (FP8) | 262k | 开放 | 20* | -- | 92 | 2.20 | 7.62 | -- | ||
Gemma 4 26B A4B | 256k | 开放 | 20* | -- | -- | -- | -- | -- | ||
Gemma 4 26B A4B | 262k | 开放 | 20* | -- | 110 | 1.09 | 5.64 | -- | ||
Gemma 4 26B A4B (FP8) | 262k | 开放 | 20* | -- | 22 | 0.73 | 23.66 | -- | ||
Gemma 4 26B A4B | 262k | 开放 | 20* | -- | 29 | 1.82 | 19.13 | -- | ||
Qwen3.5 4B (FP8) | 262k | 开放 | 20* | -- | 28 | 0.74 | 91.15 | 72.33 | ||
North Mini Code | 256k | 开放 | 20 | $0.00 | 26 | 0.58 | 95.99 | 76.33 | ||
Nova 2.0 Pro Preview (low) | 256k | 专有 | 20* | -- | 111 | 10.91 | 33.45 | 18.03 | ||
Mistral Small 4 | 256k | 开放 | 20 | $0.10 | 155 | 0.84 | 16.93 | 12.87 | ||
Devstral 2 | 262k | 开放 | 19 | $0.00 | 41 | 1.50 | 13.60 | -- | ||
Nova 2.0 Lite (medium) | 1M | 专有 | 19* | -- | 150 | 18.12 | 34.77 | 13.32 | ||
Qwen3.5 Omni Flash | 256k | 专有 | 19* | -- | 246 | 1.83 | 3.87 | -- | ||
Trinity Large Thinking | 262k | 开放 | 18* | -- | 242 | 1.48 | 11.83 | 8.28 | ||
Trinity Large Thinking (FP8) | 262k | 开放 | 18* | -- | 192 | 0.99 | 14.00 | 10.41 | ||
Magistral Medium 1.2 | 131k | 专有 | 18 | $0.89 | 69 | 2.27 | 38.34 | 28.86 | ||
Nova 2.0 Lite (low) | 1M | 专有 | 18* | -- | 150 | 9.65 | 26.28 | 13.30 | ||
Devstral Small 2 | 256k | 开放 | 18 | $0.00 | 108 | 2.33 | 6.94 | -- | ||
Qwen3 Next 80B A3B | 262k | 开放 | 17 | $0.05 | -- | -- | -- | -- | ||
Qwen3 Next 80B A3B Vertex | 262k | 开放 | 17 | $0.04 | 169 | 0.81 | 15.61 | 11.84 | ||
Qwen3 Next 80B A3B | 262k | 开放 | 17 | $0.19 | 200 | 2.32 | 14.84 | 10.01 | ||
Qwen3 Next 80B A3B (FP8) | 262k | 开放 | 17 | $0.04 | 93 | 1.19 | 28.07 | 21.50 | ||
Qwen3 Next 80B A3B | 262k | 开放 | 17 | $0.03 | -- | -- | -- | -- | ||
Nova 2.0 Omni (low) | 1M | 专有 | 17* | -- | -- | -- | -- | -- | ||
G9v3-3B | 131k | 开放 | 16* | -- | -- | -- | -- | -- | ||
Qwen3.5 4B FP8 | 262k | 开放 | 16* | -- | 22 | 0.78 | 23.35 | -- | ||
Mistral Large 3 | 256k | 开放 | 16 | $0.08 | 202 | 1.12 | 3.60 | -- | ||
Mistral Large 3 | 256k | 开放 | 16 | $0.08 | 54 | 1.90 | 11.09 | -- | ||
Mistral Large 3 | 262k | 开放 | 16 | $0.08 | 43 | 1.21 | 12.82 | -- | ||
gpt-oss-20b (high) | 131k | 开放 | 15 | $0.01 | 74 | 0.99 | 34.89 | 27.11 | ||
gpt-oss-20b (high) | 131k | 开放 | 15 | $0.03 | 140 | 36.08 | 53.91 | 14.27 | ||
gpt-oss-20b (high) | 131k | 开放 | 15 | $0.01 | 95 | 0.56 | 26.99 | 21.15 | ||
gpt-oss-20b (high) | 131k | 开放 | 15 | $0.04 | -- | -- | -- | -- | ||
gpt-oss-20b (high) | 128k | 开放 | 15 | $0.02 | 224 | 0.73 | 11.86 | 8.91 | ||
gpt-oss-20b (high) | 128k | 开放 | 15 | $0.08 | 153 | 0.89 | 17.22 | 13.06 | ||
gpt-oss-20b (high) | 131k | 开放 | 15 | $0.02 | 72 | 1.27 | 35.82 | 27.64 | ||
gpt-oss-20b (high) Vertex | 131k | 开放 | 15 | $0.03 | 282 | 3.90 | 12.77 | 7.10 | ||
gpt-oss-20b (high) | 131k | 开放 | 15 | $0.03 | 956 | 0.83 | 3.44 | 2.09 | ||
gpt-oss-20b (high) | 131k | 开放 | 15 | $0.02 | -- | -- | -- | -- | ||
Nemotron 3 Nano Omni 30B A3B | 262k | 开放 | 15* | -- | -- | -- | -- | -- | ||
Nemotron 3 Nano Omni 30B A3B (FP8) | 65.5k | 开放 | 15* | -- | 325 | 1.01 | 8.70 | 6.15 | ||
gpt-oss-120b (low) | 131k | 开放 | 15 | $0.03 | -- | -- | -- | -- | ||
gpt-oss-120b (low) | 131k | 开放 | 15 | $0.04 | -- | -- | -- | -- | ||
gpt-oss-120b (low) Base | 128k | 开放 | 15 | $0.02 | 264 | 1.04 | 10.50 | 7.57 | ||
gpt-oss-120b (low) | 131k | 开放 | 15 | $0.03 | 728 | 1.18 | 4.61 | 2.75 | ||
gpt-oss-120b (low) Vertex | 131k | 开放 | 15 | $0.01 | 184 | 1.80 | 15.40 | 10.87 | ||
gpt-oss-120b (low) | 131k | 开放 | 15 | $0.0042 | 33 | 1.53 | 78.10 | 61.25 | ||
gpt-oss-120b (low) | 131k | 开放 | 15 | $0.05 | 1,538 | 0.54 | 2.16 | 1.30 | ||
gpt-oss-120b (low) | 131k | 开放 | 15 | $0.02 | 336 | 0.85 | 8.29 | 5.95 | ||
gpt-oss-120b (low) | 128k | 开放 | 15 | $0.05 | 139 | 0.79 | 18.73 | 14.35 | ||
gpt-oss-120b (low) | 128k | 开放 | 15 | $0.01 | 198 | 0.25 | 12.87 | 10.10 | ||
gpt-oss-120b (low) | 131k | 开放 | 15 | $0.02 | 80 | 1.02 | 32.34 | 25.06 | ||
gpt-oss-120b (low) | 131k | 开放 | 15 | -- | 118 | 0.98 | 22.22 | 16.99 | ||
gpt-oss-120b (low) | 131k | 开放 | 15 | $0.02 | 110 | 0.79 | 23.42 | 18.10 | ||
gpt-oss-120b (low) | 131k | 开放 | 15 | $0.02 | 473 | 0.76 | 6.04 | 4.23 | ||
gpt-oss-120b (low) | 131k | 开放 | 15 | $0.01 | -- | -- | -- | -- | ||
gpt-oss-120b (low) | 131k | 开放 | 15 | $0.01 | 89 | 1.04 | 29.15 | 22.49 | ||
Nemotron 3 Nano | 262k | 开放 | 15 | $0.03 | 327 | 1.10 | 8.74 | 6.12 | ||
Nemotron 3 Nano (FP4) | 262k | 开放 | 15 | $0.02 | 295 | 0.99 | 9.47 | 6.78 | ||
Nemotron 3 Nano | 262k | 开放 | 15 | $0.02 | 180 | 3.25 | 17.15 | 11.12 | ||
Solar Pro 3 | 131k | 专有 | 14 | $0.15 | 133 | 2.34 | 21.07 | 14.99 | ||
Llama 4 Maverick (FP8) | 1.05M | 开放 | 14 | $0.04 | 88 | 0.81 | 6.47 | -- | ||
Llama 4 Maverick (FP8) | 128k | 开放 | 14 | $0.03 | 359 | 1.46 | 2.85 | -- | ||
Llama 4 Maverick | 128k | 开放 | 14 | $0.03 | 235 | 0.79 | 2.91 | -- | ||
Llama 4 Maverick | 131k | 开放 | 14 | $0.06 | 118 | 1.09 | 5.32 | -- | ||
Llama 4 Maverick | 1.05M | 开放 | 14 | $0.04 | -- | -- | -- | -- | ||
Llama 4 Maverick (FP8) | 1.05M | 开放 | 14 | $0.03 | 99 | 0.50 | 5.57 | -- | ||
Llama 4 Maverick (FP8) | 1.05M | 开放 | 14 | $0.03 | 115 | 1.06 | 5.42 | -- | ||
Nova 2.0 Pro Preview | 256k | 专有 | 14* | -- | 96 | 1.11 | 6.33 | -- | ||
gpt-oss-20b (low) | 131k | 开放 | 14* | -- | -- | -- | -- | -- | ||
gpt-oss-20b (low) | 131k | 开放 | 14* | -- | 938 | 0.82 | 3.48 | 2.13 | ||
gpt-oss-20b (low) | 128k | 开放 | 14* | -- | 225 | 0.73 | 11.84 | 8.89 | ||
gpt-oss-20b (low) | 131k | 开放 | 14* | -- | -- | -- | -- | -- | ||
gpt-oss-20b (low) | 131k | 开放 | 14* | -- | 72 | 0.96 | 35.59 | 27.71 | ||
gpt-oss-20b (low) | 128k | 开放 | 14* | -- | 177 | 0.82 | 14.93 | 11.29 | ||
gpt-oss-20b (low) | 131k | 开放 | 14* | -- | 172 | 14.25 | 28.79 | 11.63 | ||
gpt-oss-20b (low) | 131k | 开放 | 14* | -- | 92 | 1.19 | 28.30 | 21.68 | ||
gpt-oss-20b (low) Vertex | 131k | 开放 | 14* | -- | 178 | 8.80 | 22.84 | 11.23 | ||
Qwen3 Next 80B A3B | 262k | 开放 | 14* | -- | -- | -- | -- | -- | ||
Qwen3 Next 80B A3B Vertex | 262k | 开放 | 14* | -- | 189 | 0.80 | 3.46 | -- | ||
Qwen3 Next 80B A3B | 262k | 开放 | 14* | -- | 177 | 2.37 | 5.20 | -- | ||
Qwen3 Next 80B A3B | 262k | 开放 | 14* | -- | 98 | 0.70 | 5.82 | -- | ||
Qwen3 Next 80B A3B | 262k | 开放 | 14* | -- | 119 | 0.95 | 5.14 | -- | ||
Qwen3 Next 80B A3B | 131k | 开放 | 14* | -- | 187 | 1.63 | 4.30 | -- | ||
Qwen3 Next 80B A3B | 131k | 开放 | 14* | -- | 186 | 2.21 | 4.89 | -- | ||
Gemma 4 12B (Non-reasoning) | 262k | 开放 | 13* | -- | 112 | 2.54 | 6.99 | -- | ||
Nova Premier | 1M | 专有 | 13* | -- | 31 | 2.86 | 18.90 | -- | ||
Llama Nemotron Super 49B v1.5 | 131k | 开放 | 12* | -- | 111 | 3.97 | 26.39 | 17.94 | ||
Celeris-1 | 131k | 专有 | 12 | $0.26 | 1,560 | 0.60 | 0.92 | -- | ||
Mistral Small 4 | 262k | 开放 | 12* | -- | 138 | 0.88 | 4.49 | -- | ||
Gemma 4 E4B | 262k | 开放 | 12* | -- | 64 | 0.90 | 40.16 | 31.41 | ||
Sarvam 105B (high) | 128k | 开放 | 12* | -- | -- | -- | -- | -- | ||
Nova 2.0 Lite | 1M | 专有 | 12* | -- | 150 | 1.09 | 4.43 | -- | ||
Magistral Small 1.2 | 128k | 开放 | 11 | $0.30 | 82 | 1.51 | 32.09 | 24.47 | ||
Magistral Small 1.2 | 131k | 开放 | 11 | $0.30 | 142 | 0.85 | 18.41 | 14.05 | ||
Ministral 3 14B | 262k | 开放 | 11 | $0.16 | 96 | 0.86 | 6.07 | -- | ||
Ministral 3 14B | 256k | 开放 | 11 | $0.16 | 235 | 0.95 | 3.08 | -- | ||
Nova 2.0 Omni | 1M | 专有 | 10* | -- | -- | -- | -- | -- | ||
Llama 4 Scout | 328k | 开放 | 10 | $0.01 | 55 | 0.67 | 9.72 | -- | ||
Llama 4 Scout | 128k | 开放 | 10 | $0.01 | 129 | 0.90 | 4.77 | -- | ||
Llama 4 Scout Vertex | 1.31M | 开放 | 10 | $0.01 | 127 | 0.77 | 4.72 | -- | ||
Llama 4 Scout | 131k | 开放 | 10 | $0.02 | 62 | 0.79 | 8.83 | -- | ||
Llama 4 Scout | 128k | 开放 | 10 | $0.01 | 166 | 0.83 | 3.84 | -- | ||
Llama 4 Scout | 131k | 开放 | 10 | $0.01 | 60 | 0.90 | 9.20 | -- | ||
Llama 4 Scout | 131k | 开放 | 10 | $0.01 | 447 | 0.70 | 1.82 | -- | ||
Hermes 4 70B (FP8) | 128k | 开放 | 10* | -- | 85 | 1.38 | 30.83 | 23.56 | ||
Qwen3 Omni 30B A3B | 65.5k | 开放 | 10* | -- | 97 | 1.97 | 27.76 | 20.63 | ||
Llama 3.3 70B | 131k | 开放 | 9* | -- | -- | -- | -- | -- | ||
Llama 3.3 70B Vertex | 128k | 开放 | 9* | -- | 145 | 0.71 | 4.17 | -- | ||
Llama 3.3 70B | 131k | 开放 | 9* | -- | 291 | 0.96 | 2.68 | -- | ||
Llama 3.3 70B | 131k | 开放 | 9* | -- | 39 | 1.70 | 14.61 | -- | ||
Llama 3.3 70B (FP8) | 131k | 开放 | 9* | -- | 47 | 3.13 | 13.75 | -- | ||
Llama 3.3 70B | 128k | 开放 | 9* | -- | 286 | 2.66 | 4.41 | -- | ||
Llama 3.3 70B Base | 128k | 开放 | 9* | -- | 7 | 12.06 | 85.91 | -- | ||
Llama 3.3 70B | 128k | 开放 | 9* | -- | 76 | 0.91 | 7.51 | -- | ||
Llama 3.3 70B Turbo | 131k | 开放 | 9* | -- | 82 | 1.55 | 7.66 | -- | ||
Llama 3.3 70B | 128k | 开放 | 9* | -- | 120 | 2.25 | 6.43 | -- | ||
Llama 3.3 70B (Turbo, FP8) | 131k | 开放 | 9* | -- | 18 | 1.99 | 29.80 | -- | ||
Llama 3.3 70B | 24k | 开放 | 9* | -- | 45 | 3.77 | 15.00 | -- | ||
Llama 3.3 70B | 128k | 开放 | 9* | -- | 91 | 1.20 | 6.70 | -- | ||
Llama 3.3 70B | 100k | 开放 | 9* | -- | 82 | 1.37 | 7.45 | -- | ||
Ministral 3 8B | 256k | 开放 | 9 | $0.18 | 238 | 1.02 | 3.12 | -- | ||
Ministral 3 8B | 262k | 开放 | 9 | $0.18 | 106 | 0.75 | 5.45 | -- | ||
Llama Nemotron Ultra Base | 131k | 开放 | 9* | -- | 52 | 2.35 | 49.99 | 38.11 | ||
ERNIE 4.5 300B A47B | 131k | 开放 | 9* | -- | -- | -- | -- | -- | ||
Hermes 4 405B (FP8) | 128k | 开放 | 9* | -- | 30 | 2.56 | 87.30 | 67.79 | ||
NVIDIA Nemotron Nano 12B v2 VL (FP8) | 131k | 开放 | 9* | -- | 90 | 4.25 | 31.99 | 22.19 | ||
Gemma 4 E4B | 262k | 开放 | 9* | -- | 61 | 0.83 | 8.99 | -- | ||
NVIDIA Nemotron Nano 9B V2 | 131k | 开放 | 9* | -- | 178 | 3.78 | 17.80 | 11.22 | ||
Hermes 4 405B (FP8) | 128k | 开放 | 9* | -- | 29 | 2.47 | 19.83 | -- | ||
Llama Nemotron Super 49B v1.5 | 131k | 开放 | 9* | -- | 113 | 5.65 | 10.09 | -- | ||
LFM2.5-8B-A1B | 32.8k | 开放 | 8* | -- | 340 | 2.10 | 9.46 | 5.88 | ||
Ring-flash-2.0 | 131k | 开放 | 8* | -- | -- | -- | -- | -- | ||
Olmo 3.1 32B Think | 65.5k | 开放 | 8* | -- | -- | -- | -- | -- | ||
Command A | 288k | 开放 | 7* | -- | 58 | 1.70 | 10.33 | -- | ||
Command A | 256k | 开放 | 7* | -- | 42 | 3.29 | 15.27 | -- | ||
Llama 3.1 Nemotron 70B | 131k | 开放 | 7* | -- | 121 | 5.28 | 9.41 | -- | ||
Nemotron 3 Nano (FP4) | 262k | 开放 | 7* | -- | 301 | 0.98 | 2.64 | -- | ||
Nemotron 3 Nano | 262k | 开放 | 7* | -- | 179 | 0.39 | 3.18 | -- | ||
NVIDIA Nemotron Nano 9B V2 | 131k | 开放 | 7* | -- | 182 | 2.88 | 5.63 | -- | ||
NVIDIA Nemotron Nano 9B V2 | 131k | 开放 | 7* | -- | 165 | 1.16 | 4.20 | -- | ||
Ministral 3 3B | 256k | 开放 | 7 | $0.13 | 427 | 0.97 | 2.14 | -- | ||
Ministral 3 3B | 131k | 开放 | 7 | $0.13 | 196 | 0.65 | 3.20 | -- | ||
Hermes 4 70B (FP8) | 128k | 开放 | 7* | -- | 87 | 1.37 | 7.14 | -- | ||
Granite 4.1 8B | 131k | 开放 | 6* | -- | 113 | 0.76 | 5.17 | -- | ||
Sarvam 30B (high) | 64k | 开放 | 6* | -- | -- | -- | -- | -- | ||
Phi-4 Mini | 128k | 开放 | 6* | -- | 43 | 0.84 | 12.40 | -- | ||
Granite 4.0 H Small | 128k | 开放 | 5* | -- | 31 | 15.95 | 32.13 | -- | ||
Qwen3 Omni 30B A3B | 65.5k | 开放 | 5* | -- | 93 | 1.91 | 7.30 | -- | ||
Phi-4 | 16.4k | 开放 | 5* | -- | 35 | 2.24 | 16.48 | -- | ||
Phi-4 | 16.4k | 开放 | 5* | -- | 67 | 1.02 | 8.44 | -- | ||
Nova Micro | 130k | 专有 | 4* | -- | 277 | 0.89 | 2.70 | -- | ||
NVIDIA Nemotron Nano 12B v2 VL | 128k | 开放 | 4* | -- | 203 | 1.14 | 3.61 | -- | ||
NVIDIA Nemotron Nano 12B v2 VL (FP8) | 131k | 开放 | 4* | -- | 110 | 4.96 | 9.49 | -- | ||
Phi-4 Multimodal | 128k | 开放 | 4* | -- | 18 | 0.85 | 29.30 | -- | ||
Reka Flash 3 | 54k | 开放 | 4* | -- | -- | -- | -- | -- | ||
Llama 3.2 11B (Vision) | 131k | 开放 | 3* | -- | 20 | 1.77 | 27.10 | -- | ||
Olmo 3 7B | 65.5k | 开放 | 2* | -- | -- | -- | -- | -- | ||
Apertus 70B Instruct | 64k | 开放 | 2* | -- | -- | -- | -- | -- | ||
LFM2.5-VL-1.6B | 32.8k | 开放 | 1* | -- | 342 | 1.04 | 2.51 | -- | ||
Tiny Aya Global | 8.19k | 开放 | 1* | -- | -- | -- | -- | -- | ||
Apertus 8B Instruct | 64k | 开放 | 1* | -- | -- | -- | -- | -- | ||
Claude Sonnet 5 (low) | 1M | 专有 | -- | -- | 64 | 2.48 | 10.26 | -- | ||
Claude Sonnet 5 (xhigh) | 1M | 专有 | -- | -- | 70 | 24.34 | 31.52 | -- | ||
Claude Sonnet 5 (high) | 1M | 专有 | -- | -- | 67 | 8.10 | 15.61 | -- | ||
Cogito v2.1 | 164k | 开放 | -- | -- | -- | -- | -- | -- | ||
Claude Sonnet 5 (medium) | 1M | 专有 | -- | -- | 64 | 2.65 | 10.47 | -- | ||