模型排行
716 个模型 · 4 个榜单 · 2 个数据来源
智力王座
开源之光
性价比之王
国产之光
开源前三甲
gpt-oss-120B (high)OpenAI · 总第 12 名58.0
DeepSeek V3.1 Terminus (Reasoning)DeepSeek · 总第 13 名57.7
Qwen3 235B A22B 2507 (Reasoning)Alibaba · 总第 14 名57.5
这一期的变化
Magpie Multilingual语音榜 · 第 17上升 11 位
Octave 2语音榜 · 第 20上升 5 位
Sonic English (Oct '24)语音榜 · 第 21上升 3 位
| 排名 | 变化 | 模型 | 智力 | 价格 $/M | 速度 tok/s | 首字延迟 |
|---|---|---|---|---|---|---|
| 1 | — | GPT-5 Codex (high)闭源OpenAI | 68.5 | $3.4 | — | — |
| 2 | — | GPT-5 (high)闭源OpenAI · 另有 4 个档位 | 68.5 | $3.4 | — | — |
| 3 | — | o3闭源OpenAI | 65.5 | $3.5 | — | — |
| 4 | — | Grok 4闭源xAI | 65.3 | $6.0 | — | — |
| 5 | — | o3-pro闭源OpenAI | 65.3 | $35 | — | — |
| 6 | — | GPT-5 mini (high)闭源OpenAI · 另有 2 个档位 | 64.3 | $0.69 | — | — |
| 7 | — | Claude 4.5 Sonnet (Reasoning)闭源Anthropic · 另有 1 个档位 | 62.7 | $6.0 | — | — |
| 8 | — | Grok 4 Fast (Reasoning)闭源xAI · 另有 1 个档位 | 60.3 | $0.28 | — | — |
| 9 | — | Gemini 2.5 Pro闭源Google | 59.6 | $3.4 | — | — |
| 10 | — | o4-mini (high)闭源OpenAI | 59.6 | $1.9 | — | — |
| 11 | — | Claude 4.1 Opus (Reasoning)闭源Anthropic · 另有 1 个档位 | 59.3 | $30 | — | — |
| 12 | — | gpt-oss-120B (high)开源OpenAI | 58.0 | $0.26 | — | — |
| 13 | — | Grok 3 mini Reasoning (high)闭源xAI | 57.1 | $0.35 | — | — |
| 14 | — | Claude 4 Sonnet (Reasoning)闭源Anthropic · 另有 1 个档位 | 56.5 | $6.0 | — | — |
| 15 | — | Claude 4.5 Haiku (Reasoning)闭源Anthropic · 另有 1 个档位 | 54.6 | $2.0 | — | — |
| 16 | — | Gemini 2.5 Flash Preview (Sep '25) (Reasoning)闭源Google · 另有 3 个档位 | 54.4 | $0.85 | — | — |
| 17 | — | Claude 4 Opus (Reasoning)闭源Anthropic · 另有 1 个档位 | 54.2 | $30 | — | — |
| 18 | — | Gemini 2.5 Pro Preview (Mar' 25)闭源Google · 另有 1 个档位 | 54.1 | $3.4 | — | — |
| 19 | — | Apriel-v1.5-15B-Thinker开源ServiceNow | 51.6 | $0.000 | — | — |
| 20 | — | Gemini 2.5 Flash (Reasoning)闭源Google · 另有 1 个档位 | 51.2 | $0.85 | — | — |
| 21 | — | GPT-5 nano (high)闭源OpenAI · 另有 2 个档位 | 51.0 | $0.14 | — | — |
| 22 | — | o3-mini (high)闭源OpenAI · 另有 1 个档位 | 50.8 | $1.9 | — | — |
| 23 | — | Claude 3.7 Sonnet (Reasoning)闭源Anthropic · 另有 1 个档位 | 49.9 | $6.0 | — | — |
| 24 | — | Grok Code Fast 1闭源xAI | 48.7 | $0.53 | — | — |
| 25 | — | o1-pro闭源OpenAI | 48.0 | $263 | — | — |
| 26 | — | Gemini 2.5 Flash-Lite Preview (Sep '25) (Reasoning)闭源Google · 另有 1 个档位 | 47.9 | $0.18 | — | — |
| 27 | — | o1闭源OpenAI · 另有 1 个档位 | 47.2 | $26 | — | — |
| 28 | — | Sonar Reasoning Pro闭源Perplexity | 46.3 | $0.000 | — | — |
| 29 | — | Grok 3闭源xAI | 45.3 | $6.0 | — | — |
| 30 | — | Llama Nemotron Super 49B v1.5 (Reasoning)开源NVIDIA · 另有 1 个档位 | 45.2 | $0.18 | — | — |
| 31 | — | GPT-4.1闭源OpenAI | 43.4 | $3.5 | — | — |
| 32 | — | gpt-oss-20B (high)开源OpenAI | 43.3 | $0.095 | — | — |
| 33 | — | GPT-4.1 mini闭源OpenAI | 42.5 | $0.70 | — | — |
| 34 | — | Grok 3 Reasoning Beta闭源xAI | 41.4 | $0.000 | — | — |
| 35 | — | Gemini 2.5 Flash-Lite (Reasoning)闭源Google · 另有 1 个档位 | 40.1 | $0.18 | — | — |
| 36 | — | o1-mini闭源OpenAI | 39.2 | $1.9 | — | — |
| 37 | — | Llama 3.1 Nemotron Ultra 253B v1 (Reasoning)开源NVIDIA | 38.5 | $0.90 | — | — |
| 38 | — | GPT-4.5 (Preview)闭源OpenAI | 38.4 | $0.000 | — | — |
| 39 | — | Gemini 2.0 Flash Thinking Experimental (Jan '25)闭源Google · 另有 1 个档位 | 37.7 | $0.000 | — | — |
| 40 | — | NVIDIA Nemotron Nano 9B V2 (Reasoning)开源NVIDIA · 另有 1 个档位 | 36.9 | $0.070 | — | — |
| 41 | — | Llama 4 Maverick开源Meta | 35.8 | $0.40 | — | — |
| 42 | — | GPT-4o (March 2025, chatgpt-4o-latest)闭源OpenAI · 另有 4 个档位 | 35.6 | $7.5 | — | — |
| 43 | — | Llama 3.3 Nemotron Super 49B v1 (Reasoning)开源NVIDIA · 另有 1 个档位 | 35.5 | $0.000 | — | — |
| 44 | — | Gemini 2.0 Pro Experimental (Feb '25)闭源Google | 34.6 | $0.000 | — | — |
| 45 | — | Sonar Reasoning闭源Perplexity | 34.2 | $2.0 | — | — |
| 46 | — | Gemini 2.0 Flash (Feb '25)闭源Google · 另有 1 个档位 | 33.6 | $0.18 | — | — |
| 47 | — | Nova Premier闭源Amazon | 32.3 | $5.0 | — | — |
| 48 | — | Gemini 1.5 Pro (Sep '24)闭源Google · 另有 1 个档位 | 30.0 | $0.000 | — | — |
| 49 | — | Claude 3.5 Sonnet (Oct '24)闭源Anthropic · 另有 1 个档位 | 29.9 | $6.0 | — | — |
| 50 | — | Sonar闭源Perplexity | 28.8 | $1.0 | — | — |
智力进步
涨得最多 · Anthropic +57.2目前最高 · Anthropic 64.6最高与最低相差 18.1 分时间范围 2025年11月 – 2026年11月
64.6Anthropic60.0Google56.9OpenAI50.1Meta46.5Kimi