LLM 排行榜——比较 OpenAI、Anthropic、Google、SpaceXAI 等公司的 AI 模型
比较并排名 250 多个 AI 模型（LLM）在智能、价格、性能与速度（输出速度——每秒 token 数；延迟——TTFT）、上下文窗口等关键指标上的表现。
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。
智能
Claude Opus 5 (max) 和 Claude Opus 5 (xhigh) 是智能得分最高的模型，之后是 Claude Fable 5 (with fallback) 和 GPT-5.6 Sol (max)。
输出速度
Celeris-1 和 Mercury 2 是速度最快的模型，之后是 LFM2.5-VL-1.6B 和 Step 3.7 Flash。
延迟
Gemini 2.5 Flash-Lite 和 Command A+ 是延迟最低的模型，之后是 Gemini 2.5 Flash 和 Grok Build 0.1 0616。
每任务成本
GPT-5.6 Luna (low) 和 MiMo-V2.5 的每任务成本最低，之后是 Llama 4 Scout 和 GPT-5.6 Luna (medium)。
上下文窗口
Llama 4 Scout 和 Grok 4.20 0309 支持最大的上下文窗口，之后是 Gemini 1.5 Pro (May) 和 Grok 4.1 Fast。
进一步分析 | ||||||||
|---|---|---|---|---|---|---|---|---|
Claude Opus 5 (max) | 1M | Anthropic | 61 | $2.34 | 55 | 69.32 | 78.39 | |
Claude Opus 5 (xhigh) | 1M | Anthropic | 60 | $1.80 | 57 | 35.94 | 44.74 | |
Claude Fable 5 (with fallback) | 1M | Anthropic | 60 | $3.15 | 74 | 112.28 | 119.01 | |
GPT-5.6 Sol (max) | 1M | OpenAI | 59 | $1.23 | 66 | 139.09 | 146.69 | |
Claude Opus 5 (high) | 1M | Anthropic | 59 | $1.23 | 54 | 20.31 | 29.49 | |
GPT-5.6 Sol (xhigh) | 1M | OpenAI | 58 | $0.83 | 65 | 48.05 | 55.79 | |
Kimi K3 (max) | 1.05M |  Kimi | 57 | $0.86 | 37 | 2.87 | 71.01 | |
Claude Opus 5 (medium) | 1M | Anthropic | 56 | $0.72 | 54 | 7.49 | 16.79 | |
GPT-5.6 Sol (high) | 1M | OpenAI | 56 | $0.55 | 62 | 12.87 | 20.99 | |
GPT-5.6 Terra (max) | 1M | OpenAI | 55 | $0.51 | 122 | 171.60 | 175.69 | |
Grok 4.5 (high) | 500k | SpaceXAI | 54 | $0.36 | 59 | 11.79 | 20.20 | |
GPT-5.6 Sol (medium) | 1M | OpenAI | 54 | $0.39 | 61 | 7.89 | 16.09 | |
Claude Sonnet 5 (max) | 1M | Anthropic | 53 | $1.72 | 82 | 198.39 | 204.51 | |
GPT-5.6 Terra (xhigh) | 1M | OpenAI | 52 | $0.31 | 116 | 25.28 | 29.59 | |
GPT-5.6 Luna (max) | 1M | OpenAI | 51 | $0.05 | 178 | 135.99 | 138.80 | |
GLM-5.2 (max) | 1M |  Z AI | 51 | $0.57 | 160 | 1.52 | 17.17 | |
Muse Spark 1.1 (xhigh) | 1.05M | Meta | 51 | $0.29 | 216 | 2.90 | 14.49 | |
Claude Opus 5 (low) | 1M | Anthropic | 51 | $0.43 | 52 | 2.85 | 12.41 | |
Gemini 3.5 Flash | 1M | Google | 50 | $0.69 | 179 | 24.97 | 27.77 | |
Gemini 3.6 Flash | 1M | Google | 50 | $0.56 | 200 | 15.67 | 18.16 | |
DeepSeek V4 Flash 0731 (max) | 1M |  DeepSeek | 50 | $0.03 | 110 | 1.33 | 24.04 | |
GPT-5.6 Sol (low) | 1M | OpenAI | 49 | $0.24 | 61 | 3.74 | 11.98 | |
GPT-5.6 Luna (xhigh) | 1M | OpenAI | 49 | $0.03 | 168 | 46.39 | 49.37 | |
GPT-5.6 Terra (high) | 1M | OpenAI | 49 | $0.22 | 116 | 3.33 | 7.65 | |
Kimi K3 (low) | 1.05M | Kimi | 47 | $0.24 | 36 | 3.69 | 72.79 | |
Gemini 3.1 Pro Preview | 1M | Google | 46 | $0.34 | 126 | 32.39 | 36.35 | |
GPT-5.6 Luna (high) | 1M | OpenAI | 46 | $0.02 | 168 | 11.71 | 14.68 | |
Qwen3.7 Max | 1M |  Alibaba | 46 | $1.08 | 203 | 2.31 | 16.64 | |
GPT-5.6 Terra (medium) | 1M | OpenAI | 46 | $0.12 | 113 | 1.83 | 6.26 | |
Gemini 3.5 Flash (medium) | 1M | Google | 45* | -- | 182 | 17.04 | 19.78 | |
MiniMax-M3 | 1M |  MiniMax | 44 | $0.14 | 73 | 1.55 | 35.92 | |
DeepSeek V4 Pro (max) | 1M | DeepSeek | 44 | $0.05 | 64 | 1.61 | 77.21 | |
GPT-5.3 Codex (xhigh) | 400k | OpenAI | 44* | -- | 124 | 84.77 | 88.80 | |
Motif 3 (Beta) | 262k |  Motif Technologies | 44 | -- | -- | -- | -- | |
DeepSeek V4 Pro (high) | 1M | DeepSeek | 43 | $0.04 | 66 | 1.51 | 39.36 | |
Muse Spark | 262k | Meta | 43 | -- | -- | -- | -- | |
MiMo-V2.5-Pro | 1M | Xiaomi | 42 | $0.03 | 67 | 2.74 | 40.23 | |
Kimi K2.7 Code | 256k | Kimi | 42 | $0.22 | 42 | 2.85 | 67.31 | |
Claude Sonnet 5 (Non-reasoning) | 1M | Anthropic | 42 | $0.42 | 61 | 1.65 | 9.79 | |
Hy3 | 256k |  Tencent | 41 | $0.04 | 65 | 2.76 | 41.11 | |
GPT-5.6 Sol (Non-reasoning) | 1M | OpenAI | 41 | $0.24 | 60 | 1.05 | 9.38 | |
Nex-N2-Pro | 262k |  Nex AGI | 41 | -- | 131 | 1.60 | 20.68 | |
Inkling | 1M |  Thinking Machines | 41 | -- | 80 | 1.84 | 32.93 | |
GPT-5.6 Terra (low) | 1M | OpenAI | 40 | $0.09 | 112 | 1.64 | 6.10 | |
Inkling Small | 1M | Thinking Machines | 40 | $0.07 | 137 | 1.58 | 19.88 | |
Qwen3.6 Plus | 1M | Alibaba | 40 | $0.36 | 56 | 2.11 | 110.40 | |
Qwen3.7 Plus | 1M | Alibaba | 39 | $0.23 | 54 | 2.33 | 48.49 | |
JT-4.1 Flash 236B A21B | 256k |  China Mobile | 39 | -- | -- | -- | -- | |
Agnes 2.5 Pro Alpha | 1M |  Sapiens AI | 39 | -- | 134 | 2.72 | 21.40 | |
GPT-5.6 Luna (medium) | 1M | OpenAI | 38 | $0.01 | 151 | 3.01 | 6.33 | |
Nemotron 3 Ultra | 262k |  NVIDIA | 38 | $0.38 | 145 | 2.15 | 21.23 | |
MiMo-V2.5 | 1M | Xiaomi | 37 | $0.01 | 85 | 2.56 | 31.94 | |
Qwen3.6 27B | 262k | Alibaba | 37 | $0.29 | 59 | 3.77 | 108.76 | |
Gemini 3.5 Flash-Lite | 1M | Google | 36 | $0.10 | 346 | 10.30 | 11.75 | |
MiMo-V2-Omni-0327 | 256k | Xiaomi | 36* | -- | -- | -- | -- | |
Grok 4.3 (medium) | 1M | SpaceXAI | 36* | -- | 127 | 11.57 | 15.52 | |
Grok 4.3 (low) | 1M | SpaceXAI | 35* | -- | 112 | 6.49 | 10.98 | |
MiMo-V2-Omni | 256k | Xiaomi | 35* | -- | -- | -- | -- | |
Gemini 3.5 Flash (minimal) | 1M | Google | 35* | -- | 171 | 1.05 | 3.97 | |
Kimi K2.6 | 256k | Kimi | 35* | -- | 35 | 2.89 | 17.02 | |
Claude Sonnet 4.6 (Non-reasoning, Low Effort) | 1M | Anthropic | 34* | -- | 44 | 1.72 | 13.11 | |
GLM-5.2 | 1M | Z AI | 34 | -- | 97 | 1.85 | 7.00 | |
GPT-5.6 Terra (Non-reasoning) | 1M | OpenAI | 34 | $0.10 | 107 | 0.79 | 5.48 | |
KAT-Coder-Pro V2 | 256k |  KwaiKAT | 34 | -- | 108 | 1.87 | 6.50 | |
Qwen3.5 397B A17B | 262k | Alibaba | 34 | $0.36 | 70 | 2.37 | 54.83 | |
Hy3-preview | 256k | Tencent | 34* | -- | -- | -- | -- | |
LongCat 2.0 | 1M |  LongCat | 33 | $0.12 | 42 | 2.74 | 61.75 | |
GPT-5.6 Luna (low) | 1M | OpenAI | 33 | $0.01 | 137 | 2.12 | 5.77 | |
MiMo-V2-Flash (Feb 2026) | 256k | Xiaomi | 33* | -- | -- | -- | -- | |
Qwen3.5 122B A10B | 262k | Alibaba | 32 | $0.26 | 125 | 2.57 | 22.64 | |
Qwen3.5 397B A17B | 262k | Alibaba | 32* | -- | 72 | 2.44 | 9.42 | |
Qwen3.6 35B A3B | 262k | Alibaba | 32 | $0.19 | 141 | 2.30 | 43.97 | |
DeepSeek V4 Pro | 1M | DeepSeek | 31* | -- | 63 | 1.52 | 9.41 | |
G9v3-39A5B | 131k |  AI9Stars | 31 | -- | -- | -- | -- | |
Qwen3.5 Omni Plus | 256k | Alibaba | 31* | -- | 53 | 2.36 | 11.83 | |
Ring-2.6-1T | 262k |  InclusionAI | 31 | $0.37 | -- | -- | -- | |
Qwen3.6 27B | 262k | Alibaba | 30 | $0.40 | 57 | 3.72 | 12.50 | |
o3 | 200k | OpenAI | 30* | -- | 120 | 8.41 | 12.57 | |
Step 3.7 Flash | 262k | StepFun | 30 | $0.09 | 364 | 0.92 | 7.79 | |
Mistral Medium 3.5 | 256k |  Mistral | 30 | $0.46 | 134 | 1.88 | 20.57 | |
Claude 4.5 Haiku | 200k | Anthropic | 30 | $0.22 | 99 | 18.99 | 24.05 | |
Gemma 4 31B | 256k | Google | 29 | $0.00 | 35 | 1.10 | 64.26 | |
GPT-5.5 Instant (June 2026) | 400k | OpenAI | 29 | $0.54 | 168 | 1.00 | 15.93 | |
DeepSeek V4 Flash | 1M | DeepSeek | 29* | -- | 120 | 1.20 | 5.36 | |
JT-35B-Flash | 256k | China Mobile | 28* | -- | -- | -- | -- | |
KAT-Coder-Pro V1 | 256k | KwaiKAT | 28* | -- | -- | -- | -- | |
MiMo-V2.5-Pro | 1M | Xiaomi | 28* | -- | 61 | 3.24 | 11.37 | |
Qwen3.5 122B A10B | 262k | Alibaba | 28 | $0.19 | 140 | 2.56 | 6.12 | |
GPT-5.6 Luna (Non-reasoning) | 1M | OpenAI | 27 | $0.01 | 162 | 0.82 | 3.91 | |
Hy3-preview | 256k | Tencent | 26* | -- | -- | -- | -- | |
Ling-2.6-1T | 262k | InclusionAI | 26* | -- | -- | -- | -- | |
Doubao Seed Code | 256k |  ByteDance Seed | 26* | -- | -- | -- | -- | |
Gemini 2.5 Pro | 1M | Google | 26 | $0.22 | 136 | 22.36 | 26.03 | |
Gemma 4 26B A4B | 256k | Google | 26 | $0.04 | -- | -- | -- | |
NVIDIA Nemotron 3 Super | 1M | NVIDIA | 25 | $0.23 | 143 | 1.95 | 19.40 | |
Gemini 3.1 Flash-Lite | 1M | Google | 25 | $0.04 | 312 | 6.13 | 7.73 | |
Grok 4.3 (Non-reasoning) | 1M | SpaceXAI | 25 | $0.29 | 120 | 0.73 | 4.91 | |
MiMo-V2-Flash | 256k | Xiaomi | 25 | -- | -- | -- | -- | |
Qwen3.6 35B A3B | 262k | Alibaba | 24 | $0.63 | 164 | 2.42 | 5.47 | |
Qwen3.5 35B A3B | 262k | Alibaba | 24 | $0.24 | 170 | 2.35 | 5.29 | |
gpt-oss-120b (high) | 131k | OpenAI | 24 | $0.08 | 185 | 0.85 | 14.38 | |
Claude 4.5 Haiku | 200k | Anthropic | 24* | -- | 90 | 0.98 | 6.54 | |
Command A+ | 192k | Cohere | 23 | $0.00 | 209 | 0.44 | 12.38 | |
K-EXAONE | 256k |  LG AI Research | 22 | -- | -- | -- | -- | |
ERNIE 5.0 Thinking Preview | 128k |  Baidu | 22* | -- | -- | -- | -- | |
Gemma 4 12B | 256k | Google | 22 | $0.08 | 111 | 2.39 | 24.86 | |
Gemma 4 31B | 256k | Google | 22 | $0.04 | 66 | 2.44 | 9.98 | |
Nova 2.0 Pro Preview (medium) | 256k |  Amazon | 22 | $0.18 | 118 | 14.71 | 35.85 | |
Qwen3.5 9B | 262k | Alibaba | 21 | $0.24 | 73 | 1.66 | 35.97 | |
Mercury 2 | 128k | Inception | 21 | $0.08 | 742 | 3.28 | 3.95 | |
Qwen3 Coder Next | 256k | Alibaba | 21 | $0.33 | 133 | 1.71 | 5.45 | |
Nova 2.0 Omni (medium) | 1M | Amazon | 21* | -- | -- | -- | -- | |
Apriel-v1.6-15B-Thinker | 128k |  ServiceNow | 21* | -- | -- | -- | -- | |
Qwen3.5 9B | 262k | Alibaba | 20* | -- | -- | -- | -- | |
EXAONE 4.5 33B | 262k | LG AI Research | 20 | -- | -- | -- | -- | |
Gemma 4 26B A4B | 256k | Google | 20* | -- | 79 | 1.31 | 7.67 | |
Qwen3.5 4B | 262k | Alibaba | 20* | -- | 32 | 0.67 | 78.55 | |
North Mini Code | 256k | Cohere | 20 | $0.00 | 21 | 0.62 | 117.18 | |
Nova 2.0 Pro Preview (low) | 256k | Amazon | 20 | $0.23 | 123 | 11.68 | 32.00 | |
Mistral Small 4 | 256k | Mistral | 20 | $0.10 | 158 | 1.00 | 16.83 | |
Devstral 2 | 256k | Mistral | 19 | $0.00 | 43 | 1.47 | 13.16 | |
Nova 2.0 Lite (medium) | 1M | Amazon | 19* | -- | 155 | 17.81 | 33.96 | |
Qwen3.5 Omni Flash | 256k | Alibaba | 19* | -- | 242 | 1.86 | 3.93 | |
JT-MINI | 128k | China Mobile | 19* | -- | -- | -- | -- | |
Nova 2.0 Lite (high) | 1M | Amazon | 18 | $0.24 | 138 | 16.25 | 34.32 | |
Trinity Large Thinking | 512k |  Arcee AI | 18 | $0.16 | 205 | 1.03 | 13.21 | |
Magistral Medium 1.2 | 128k | Mistral | 18 | $0.84 | 110 | 2.17 | 24.96 | |
Nova 2.0 Lite (low) | 1M | Amazon | 18* | -- | 156 | 10.57 | 26.55 | |
HyperNova 60B 2605 | 131k |  Multiverse Computing | 18 | $0.02 | 349 | 0.93 | 8.10 | |
Nemotron Cascade 2 30B A3B | 1M | NVIDIA | 18 | -- | -- | -- | -- | |
Devstral Small 2 | 256k | Mistral | 17 | $0.00 | 89 | 1.74 | 7.38 | |
K2 Think V2 | 262k |  MBZUAI Institute of Foundation Models | 17 | -- | -- | -- | -- | |
LongCat Flash Lite | 256k | LongCat | 17* | -- | -- | -- | -- | |
HyperCLOVA X SEED Think (32B) | 128k |  Naver | 17* | -- | -- | -- | -- | |
K-EXAONE | 256k | LG AI Research | 17* | -- | -- | -- | -- | |
Qwen3 Next 80B A3B | 262k | Alibaba | 17 | $0.19 | 205 | 2.41 | 14.61 | |
Nova 2.0 Omni (low) | 1M | Amazon | 17* | -- | -- | -- | -- | |
Mi:dm K 2.5 Pro | 128k |  Korea Telecom | 16* | -- | -- | -- | -- | |
G9v3-3B | 131k | AI9Stars | 16 | $0.00 | -- | -- | -- | |
Qwen3.5 4B | 262k | Alibaba | 16* | -- | 19 | 0.96 | 26.91 | |
Mistral Large 3 | 256k | Mistral | 16 | $0.08 | 43 | 1.22 | 12.88 | |
INTELLECT-3 | 131k |  Prime Intellect | 16* | -- | -- | -- | -- | |
Solar Open 100B | 128k |  Upstage | 15* | -- | -- | -- | -- | |
Nemotron 3 Nano Omni 30B A3B Reasoning | 256k | NVIDIA | 15* | -- | 324 | 1.01 | 8.73 | |
gpt-oss-120b (low) | 131k | OpenAI | 15 | $0.02 | 244 | 0.89 | 11.12 | |
gpt-oss-20b (high) | 131k | OpenAI | 15 | $0.03 | 166 | 0.89 | 15.96 | |
Nova 2.0 Pro Preview | 256k | Amazon | 14 | $0.26 | 96 | 1.01 | 6.24 | |
gpt-oss-20b (low) | 131k | OpenAI | 14* | -- | 180 | 0.93 | 14.83 | |
Llama 4 Maverick | 1M | Meta | 14 | $0.04 | 106 | 0.92 | 5.62 | |
K2-V2 (high) | 512k | MBZUAI Institute of Foundation Models | 14* | -- | -- | -- | -- | |
NVIDIA Nemotron 3 Nano | 1M | NVIDIA | 14 | $0.02 | 248 | 1.47 | 11.55 | |
Solar Pro 3 | 128k | Upstage | 14 | $0.14 | 125 | 2.29 | 22.35 | |
Ling 2.6 Flash | 262k | InclusionAI | 14 | -- | 100 | 1.17 | 6.17 | |
Qwen3 Next 80B A3B | 262k | Alibaba | 14* | -- | 185 | 2.17 | 4.88 | |
DiffusionGemma 26B A4B | 256k | Google | 13 | -- | -- | -- | -- | |
Gemma 4 12B (Non-reasoning) | 262k | Google | 13* | -- | 108 | 2.38 | 7.01 | |
Motif-2-12.7B | 128k | Motif Technologies | 13* | -- | -- | -- | -- | |
Nova Premier | 1M | Amazon | 13* | -- | 32 | 2.85 | 18.65 | |
K2-V2 (medium) | 512k | MBZUAI Institute of Foundation Models | 12* | -- | -- | -- | -- | |
Llama Nemotron Super 49B v1.5 | 128k | NVIDIA | 12* | -- | 81 | 5.85 | 36.73 | |
Mistral Small 4 | 256k | Mistral | 12* | -- | 130 | 0.93 | 4.76 | |
Tri-21B-Think | 32k |  Trillion Labs | 12* | -- | -- | -- | -- | |
Gemma 4 E4B | 128k | Google | 12 | $0.01 | 82 | 0.81 | 31.19 | |
MiniCPM5-1B | 128k |  OpenBMB | 12* | -- | -- | -- | -- | |
Sarvam 105B (high) | 128k |  Sarvam | 12* | -- | -- | -- | -- | |
Nova 2.0 Lite | 1M | Amazon | 12* | -- | 152 | 1.08 | 4.37 | |
Celeris-1 | 131k | Celeris | 12 | -- | 2,034 | 0.69 | 0.94 | |
MiniCPM5-1B | 128k | OpenBMB | 12* | -- | -- | -- | -- | |
Magistral Small 1.2 | 128k | Mistral | 11 | $0.29 | 137 | 0.84 | 19.10 | |
Nanbeige4.1-3B | 256k |  Nanbeige | 11 | -- | -- | -- | -- | |
Ministral 3 14B | 256k | Mistral | 11 | $0.16 | 95 | 0.87 | 6.14 | |
EXAONE 4.0 32B | 131k | LG AI Research | 11* | -- | -- | -- | -- | |
Nova 2.0 Omni | 1M | Amazon | 11* | -- | -- | -- | -- | |
Llama 4 Scout | 10M | Meta | 10 | $0.01 | 126 | 0.81 | 4.79 | |
Hermes 4 70B | 128k |  Nous Research | 10* | -- | 89 | 1.39 | 29.41 | |
Falcon-H1R-7B | 256k |  TII UAE | 10* | -- | -- | -- | -- | |
Gemma 4 E2B | 128k | Google | 10 | -- | -- | -- | -- | |
Qwen3 Omni 30B A3B | 65.5k | Alibaba | 10* | -- | 102 | 1.99 | 26.62 | |
Step3 VL 10B | 65.5k | StepFun | 9* | -- | -- | -- | -- | |
Llama 3.3 70B | 128k | Meta | 9 | $0.09 | 81 | 1.56 | 7.72 | |
Llama Nemotron Ultra | 128k | NVIDIA | 9* | -- | 52 | 2.34 | 50.08 | |
ERNIE 4.5 300B A47B | 131k | Baidu | 9* | -- | -- | -- | -- | |
Hermes 4 405B | 128k | Nous Research | 9* | -- | 32 | 2.63 | 81.09 | |
NVIDIA Nemotron Nano 12B v2 VL | 128k | NVIDIA | 9* | -- | 86 | 7.19 | 36.26 | |
Ministral 3 8B | 256k | Mistral | 9 | $0.18 | 118 | 0.73 | 4.97 | |
Gemma 4 E4B | 128k | Google | 9* | -- | 82 | 0.80 | 6.86 | |
Granite 4.1 30B | 131k |  IBM | 9 | -- | -- | -- | -- | |
NVIDIA Nemotron Nano 9B V2 | 131k | NVIDIA | 9* | -- | 180 | 5.20 | 19.08 | |
Hermes 4 405B | 128k | Nous Research | 9* | -- | 35 | 2.36 | 16.45 | |
NVIDIA Nemotron 3 Nano 4B | 262k | NVIDIA | 9 | -- | -- | -- | -- | |
Llama Nemotron Super 49B v1.5 | 128k | NVIDIA | 9* | -- | 98 | 6.45 | 11.53 | |
K2-V2 (low) | 512k | MBZUAI Institute of Foundation Models | 9* | -- | -- | -- | -- | |
Kimi Linear 48B A3B Instruct | 1M | Kimi | 9* | -- | -- | -- | -- | |
Llama 3.1 405B | 128k | Meta | 9* | -- | -- | -- | -- | |
LFM2.5-8B-A1B | 32.8k | Liquid AI | 8* | -- | 338 | 1.46 | 8.85 | |
Ring-flash-2.0 | 128k | InclusionAI | 8* | -- | -- | -- | -- | |
Olmo 3.1 32B Think | 65.5k |  Allen Institute for AI | 8* | -- | -- | -- | -- | |
Command A | 256k | Cohere | 8* | -- | 59 | 1.84 | 10.32 | |
Qwen3.5 2B | 262k | Alibaba | 8 | -- | -- | -- | -- | |
Llama 3.1 Nemotron 70B | 128k | NVIDIA | 8* | -- | 116 | 6.63 | 10.95 | |
NVIDIA Nemotron 3 Nano | 1M | NVIDIA | 7* | -- | 192 | 0.98 | 3.58 | |
NVIDIA Nemotron Nano 9B V2 | 131k | NVIDIA | 7* | -- | 171 | 1.41 | 4.34 | |
Hermes 4 70B | 128k | Nous Research | 7* | -- | 93 | 1.36 | 6.74 | |
Ministral 3 3B | 256k | Mistral | 7 | $0.13 | 206 | 0.66 | 3.09 | |
Granite 4.1 8B | 131k | IBM | 7* | -- | 103 | 0.79 | 5.62 | |
Sarvam 30B (high) | 65.5k | Sarvam | 7* | -- | -- | -- | -- | |
Olmo 3.1 32B Instruct | 65.5k | Allen Institute for AI | 6* | -- | -- | -- | -- | |
Gemma 4 E2B | 128k | Google | 6* | -- | -- | -- | -- | |
R1 1776 | 128k |  Perplexity | 6* | -- | -- | -- | -- | |
Llama 3.2 90B (Vision) | 128k | Meta | 6* | -- | -- | -- | -- | |
Phi-4 Mini | 128k |  Microsoft | 6 | $0.00 | 43 | 0.84 | 12.53 | |
EXAONE 4.0 32B | 131k | LG AI Research | 6* | -- | -- | -- | -- | |
Qwen3.5 2B | 262k | Alibaba | 6 | -- | -- | -- | -- | |
Qwen3.5 0.8B | 262k | Alibaba | 5 | -- | -- | -- | -- | |
DeepHermes 3 - Mistral 24B | 32k | Nous Research | 5* | -- | -- | -- | -- | |
Jamba 1.7 Large | 256k |  AI21 Labs | 5* | -- | 55 | 1.42 | 10.49 | |
Granite 4.0 H Small | 128k | IBM | 5* | -- | 46 | 10.49 | 21.45 | |
Qwen3 Omni 30B A3B | 65.5k | Alibaba | 5* | -- | 95 | 1.89 | 7.13 | |
LFM2 24B A2B | 32.8k | Liquid AI | 5* | -- | -- | -- | -- | |
Phi-4 | 16k | Microsoft | 5* | -- | -- | -- | -- | |
Nova Micro | 130k | Amazon | 5* | -- | 282 | 0.86 | 2.63 | |
Granite 4.1 3B | 131k | IBM | 5 | -- | -- | -- | -- | |
NVIDIA Nemotron Nano 12B v2 VL | 128k | NVIDIA | 5* | -- | 182 | 1.68 | 4.43 | |
Phi-4 Multimodal | 128k | Microsoft | 5* | -- | 18 | 0.81 | 29.09 | |
MiniCPM-V 4.6 1.3B | 262k | OpenBMB | 4 | -- | -- | -- | -- | |
Jamba Reasoning 3B | 262k | AI21 Labs | 4* | -- | -- | -- | -- | |
Reka Flash 3 | 128k |  Reka AI | 4* | -- | -- | -- | -- | |
Olmo 3 7B Think | 65.5k | Allen Institute for AI | 4* | -- | -- | -- | -- | |
Molmo 7B-D | 4.1k | Allen Institute for AI | 4* | -- | -- | -- | -- | |
Ling-mini-2.0 | 131k | InclusionAI | 4* | -- | -- | -- | -- | |
Llama 3.2 11B (Vision) | 128k | Meta | 3* | -- | 37 | 1.70 | 15.08 | |
Qwen3.5 0.8B | 262k | Alibaba | 3 | -- | -- | -- | -- | |
Exaone 4.0 1.2B | 64k | LG AI Research | 3* | -- | -- | -- | -- | |
Olmo 3 7B | 65.5k | Allen Institute for AI | 3* | -- | -- | -- | -- | |
Exaone 4.0 1.2B | 64k | LG AI Research | 3* | -- | -- | -- | -- | |
LFM2.5-1.2B-Thinking | 32k | Liquid AI | 3* | -- | -- | -- | -- | |
Jamba 1.7 Mini | 258k | AI21 Labs | 3* | -- | -- | -- | -- | |
LFM2 2.6B | 32.8k | Liquid AI | 3* | -- | -- | -- | -- | |
LFM2.5-1.2B-Instruct | 32k | Liquid AI | 3* | -- | -- | -- | -- | |
Granite 4.0 H 1B | 128k | IBM | 3* | -- | -- | -- | -- | |
Gemma 3 270M | 32k | Google | 2* | -- | -- | -- | -- | |
Apertus 70B Instruct | 65.5k |  Swiss AI Initiative | 2* | -- | -- | -- | -- | |
Granite 4.0 Micro | 128k | IBM | 2* | -- | -- | -- | -- | |
DeepHermes 3 - Llama-3.1 8B | 128k | Nous Research | 2* | -- | -- | -- | -- | |
Granite 4.0 1B | 128k | IBM | 2* | -- | -- | -- | -- | |
Molmo2-8B | 36.9k | Allen Institute for AI | 2* | -- | -- | -- | -- | |
LFM2 8B A1B | 32.8k | Liquid AI | 2* | -- | -- | -- | -- | |
LFM2.5-VL-1.6B | 32k | Liquid AI | 1* | -- | 474 | 1.25 | 2.30 | |
Granite 4.0 350M | 32.8k | IBM | 1* | -- | -- | -- | -- | |
Tiny Aya Global | 8.19k | Cohere | 1* | -- | -- | -- | -- | |
Apertus 8B Instruct | 65.5k | Swiss AI Initiative | 1* | -- | -- | -- | -- | |
Granite 4.0 H 350M | 32.8k | IBM | 1* | -- | -- | -- | -- | |
Claude Sonnet 5 (low) | 1M | Anthropic | -- | -- | 61 | 1.68 | 9.88 | |
EXAONE 4.5 33B | 262k | LG AI Research | -- | -- | -- | -- | -- | |
Claude Sonnet 5 (xhigh) | 1M | Anthropic | -- | -- | 71 | 36.78 | 43.78 | |
Gemini 3 Deep Think | 128k | Google | -- | -- | -- | -- | -- | |
Claude Sonnet 5 (high) | 1M | Anthropic | -- | -- | 68 | 10.56 | 17.96 | |
Mi:dm K 2.5 Pro Preview | 128k | Korea Telecom | -- | -- | -- | -- | -- | |
GPT-5.5 Pro (xhigh) | 922k | OpenAI | -- | -- | -- | -- | -- | |
Cogito v2.1 | 128k |  Deep Cogito | -- | -- | -- | -- | -- | |
Claude Sonnet 5 (medium) | 1M | Anthropic | -- | -- | 64 | 3.64 | 11.50 | |
关键定义
常见问题
在参与排名的 175 个模型中，Claude Opus 5 (Adaptive Reasoning, Max Effort) 目前以 61 的 Intelligence Index 得分位居 Artificial Analysis LLM 排行榜第一。
按 Intelligence Index 排名，顶尖模型是：1. Claude Opus 5 (Adaptive Reasoning, Max Effort)（61）, 2. Claude Opus 5 (Adaptive Reasoning, Xhigh Effort)（60）, 3. Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)（60）, 4. GPT-5.6 Sol (max)（59）, 5. Claude Opus 5 (Adaptive Reasoning, High Effort)（59）。
Celeris-1 最快，输出速度为每秒 2,033.7 个 token，其次是 Mercury 2（741.8 t/s）和 LFM2.5-VL-1.6B（474.5 t/s）。
GPT-5.6 Luna (low) 的每项 Intelligence Index 任务成本最低，为 $0.01，其次是 MiMo-V2.5（$0.01）和 Llama 4 Scout（$0.01）。
Kimi K3 (max) 是排名最高的开放权重模型，Intelligence Index 得分为 57。排行榜上的 175 个模型中，有 99 个开放权重模型。
按 Intelligence Index 排名，顶尖开放权重模型是：1. Kimi K3 (max)（57）, 2. GLM-5.2 (max)（51）, 3. DeepSeek V4 Flash 0731 (Reasoning, Max Effort)（50）。
在 130 个推理模型中，Claude Opus 5 (Adaptive Reasoning, Max Effort) 以 61 的 Intelligence Index 得分领先。推理模型会先通过扩展思考来解决复杂问题，再给出回答。
排行榜提供筛选条件，可按模型类型（推理与非推理）、开放性（开放权重与专有）及其他标准缩小结果范围。您还可以调整提示词选项，查看不同输入长度下的性能变化。
点击排行榜中的任意模型名称，即可访问其专属比较页面，查看智能、价格、速度、延迟等详细图表。您还可以比较每个模型的 API 服务商。 查看所有模型