模型排行
569 個模型 · 3 個榜單 · 2 個數據來源
人氣王座
開源之光
國產之光
最多人測
得分前三甲
claude-sonnet-4-5-20250929-thinking-32kAnthropic1,453
gemini-2.5-proGoogle1,452
claude-opus-4-1-20250805-thinking-16kAnthropic1,449
開源前三甲
glm-4.6Z.ai · 總第 12 名1,426
qwen3-235b-a22b-instruct-2507Alibaba · 總第 16 名1,419
deepseek-v3.2-expDeepSeek · 總第 17 名1,418
這一期的變化
GPT-5 Codex (high)智力榜 · 第 2新上榜
Claude 4.5 Sonnet (Reasoning)智力榜 · 第 7新上榜
DeepSeek V3.1 Terminus (Reasoning)智力榜 · 第 13新上榜
Gemini 2.5 Flash-Lite Preview (Sep '25) (Reasoning)智力榜 · 第 34新上榜
Qwen3 235B A22B 2507 Instruct智力榜 · 第 39新上榜
| 排名 | 變化 | 模型 | 得分 | 置信區間 | 投票 | 價格 $/M |
|---|---|---|---|---|---|---|
| 1 | 新 | glm-4.6開源Z.ai | 1,426 | ±13 | 2,045 | — |
| 2 | ▼ 1 | qwen3-235b-a22b-instruct-2507開源Alibaba | 1,419 | ±5 | 26,731 | — |
| 3 | 新 | deepseek-v3.2-exp開源DeepSeek | 1,418 | ±13 | 2,100 | — |
| 4 | 新 | qwen3-vl-235b-a22b-instruct開源Alibaba | 1,418 | ±9 | 3,950 | — |
| 5 | ▼ 2 | deepseek-r1-0528開源DeepSeek · 另有 1 個檔位 | 1,417 | ±6 | 19,488 | — |
| 6 | ▼ 4 | kimi-k2-0905-preview開源Moonshot AI · 另有 1 個檔位 | 1,416 | ±7 | 10,935 | — |
| 7 | ▼ 3 | deepseek-v3.1-thinking開源DeepSeek · 另有 1 個檔位 | 1,416 | ±7 | 12,282 | — |
| 8 | 新 | deepseek-v3.1-terminus開源DeepSeek · 另有 1 個檔位 | 1,415 | ±10 | 3,834 | — |
| 9 | ▼ 4 | glm-4.5開源Z.ai | 1,406 | ±5 | 20,602 | — |
| 10 | ▼ 4 | qwen3-next-80b-a3b-instruct開源Alibaba | 1,403 | ±7 | 10,238 | — |
| 11 | ▼ 2 | longcat-flash-chat開源Meituan | 1,399 | ±7 | 9,946 | — |
| 12 | ▼ 5 | qwen3-235b-a22b-no-thinking開源Alibaba | 1,397 | ±5 | 38,121 | — |
| 13 | ▼ 5 | qwen3-235b-a22b-thinking-2507開源Alibaba | 1,397 | ±6 | 9,490 | — |
| 14 | ▼ 4 | deepseek-v3-0324開源DeepSeek · 另有 1 個檔位 | 1,390 | ±4 | 42,219 | — |
| 15 | 新 | qwen3-vl-235b-a22b-thinking開源Alibaba | 1,387 | ±10 | 3,712 | — |
| 16 | ▼ 5 | qwen3-30b-a3b-instruct-2507開源Alibaba | 1,384 | ±5 | 19,470 | — |
| 17 | ▼ 5 | qwen3-coder-480b-a35b-instruct開源Alibaba | 1,382 | ±5 | 21,683 | — |
| 18 | ▼ 4 | qwen3-235b-a22b開源Alibaba | 1,372 | ±5 | 25,727 | — |
| 19 | ▼ 6 | qwen3-next-80b-a3b-thinking開源Alibaba | 1,372 | ±7 | 9,116 | — |
| 20 | ▼ 5 | glm-4.5-air開源Z.ai | 1,369 | ±5 | 19,610 | — |
| 21 | ▼ 5 | minimax-m1開源MiniMax | 1,368 | ±5 | 29,909 | — |
| 22 | ▼ 5 | gemma-3-27b-it開源Google | 1,362 | ±4 | 42,372 | — |
| 23 | ▼ 5 | mistral-small-2506開源Mistral | 1,351 | ±5 | 16,805 | — |
| 24 | 新 | glm-4.5v開源Z.ai | 1,349 | ±10 | 3,295 | — |
| 25 | ▼ 6 | gpt-oss-120b開源OpenAI | 1,347 | ±6 | 18,789 | $0.26 |
| 26 | ▼ 5 | command-a-03-2025開源Cohere | 1,347 | ±4 | 45,172 | — |
| 27 | ▼ 5 | llama-3.1-nemotron-ultra-253b-v1開源NVIDIA | 1,344 | ±12 | 2,573 | — |
| 28 | ▼ 8 | step-3開源StepFun | 1,344 | ±8 | 5,050 | — |
| 29 | ▼ 6 | qwen3-32b開源Alibaba | 1,343 | ±10 | 3,943 | — |
| 30 | ▼ 5 | gemma-3-12b-it開源Google | 1,340 | ±10 | 3,866 | — |
| 31 | ▼ 7 | nvidia-llama-3.3-nemotron-super-49b-v1.5開源NVIDIA | 1,339 | ±10 | 3,548 | — |
| 32 | ▼ 5 | llama-3.1-405b-instruct-bf16開源Meta | 1,334 | ±4 | 41,932 | — |
| 33 | ▼ 7 | qwq-32b開源Alibaba · 另有 1 個檔位 | 1,333 | ±5 | 25,118 | — |
| 34 | ▼ 6 | llama-3.1-405b-instruct-fp8開源Meta | 1,332 | ±3 | 60,272 | — |
| 35 | ▼ 4 | llama-4-maverick-17b-128e-instruct開源Meta | 1,326 | ±4 | 39,766 | — |
| 36 | ▼ 7 | qwen3-30b-a3b開源Alibaba | 1,324 | ±5 | 25,917 | — |
| 37 | ▼ 5 | llama-3.3-nemotron-49b-super-v1開源NVIDIA | 1,324 | ±12 | 2,243 | — |
| 38 | ▼ 4 | llama-4-scout-17b-16e-instruct開源Meta | 1,321 | ±5 | 29,809 | — |
| 39 | ▼ 6 | deepseek-v2.5-1210開源DeepSeek · 另有 1 個檔位 | 1,320 | ±8 | 6,877 | — |
| 40 | ▼ 10 | gpt-oss-20b開源OpenAI | 1,320 | ±7 | 9,300 | — |
| 41 | ▼ 5 | llama-3.3-70b-instruct開源Meta | 1,318 | ±3 | 54,258 | — |
| 42 | ▼ 7 | gemma-3n-e4b-it開源Google | 1,318 | ±5 | 23,761 | — |
| 43 | ▼ 6 | qwen-max-0919開源Alibaba | 1,316 | ±6 | 16,598 | — |
| 44 | ▼ 5 | mistral-large-2407開源Mistral · 另有 1 個檔位 | 1,312 | ±4 | 45,968 | — |
| 45 | ▼ 7 | athene-v2-chat開源NexusFlow | 1,312 | ±4 | 24,880 | — |
| 46 | ▼ 6 | athene-70b-0725開源NexusFlow | 1,303 | ±6 | 19,796 | — |
| 47 | ▼ 6 | gemma-3-4b-it開源Google | 1,302 | ±9 | 4,195 | — |
| 48 | ▼ 6 | qwen2.5-72b-instruct開源Alibaba | 1,300 | ±4 | 39,632 | — |
| 49 | ▼ 6 | mistral-small-3.1-24b-instruct-2503開源Mistral | 1,298 | ±5 | 29,513 | — |
| 50 | ▼ 6 | llama-3.1-nemotron-70b-instruct開源NVIDIA | 1,296 | ±8 | 7,216 | — |
智力進步
漲得最多 · Anthropic +57.2目前最高 · Anthropic 64.6最高與最低相差 18.1 分時間範圍 2025年11月 – 2026年11月
64.6Anthropic60.0Google56.9OpenAI50.1Meta46.5Kimi