模型排行
165 个模型 · 2 个榜单 · 1 个数据来源
人气王座
开源之光
国产之光
最多人测
开源前三甲
deepseek-v3DeepSeek · 总第 4 名1,279
llama-3.1-405b-instruct-bf16Meta · 总第 9 名1,254
llama-3.1-405b-instruct-fp8Meta · 总第 11 名1,253
| 排名 | 变化 | 模型 | 得分 | 置信区间 | 投票 | 价格 $/M |
|---|---|---|---|---|---|---|
| 1 | — | deepseek-v3开源DeepSeek | 1,279 | ±5 | 11,591 | — |
| 2 | ▲ 1 | llama-3.1-405b-instruct-bf16开源Meta | 1,254 | ±5 | 21,285 | — |
| 3 | ▼ 1 | llama-3.1-405b-instruct-fp8开源Meta | 1,253 | ±3 | 63,202 | — |
| 4 | — | deepseek-v2.5-1210开源DeepSeek · 另有 1 个档位 | 1,240 | ±6 | 7,261 | — |
| 5 | — | llama-3.3-70b-instruct开源Meta | 1,235 | ±4 | 15,516 | — |
| 6 | — | athene-v2-chat开源NexusFlow | 1,232 | ±4 | 21,014 | — |
| 7 | — | mistral-large-2407开源Mistral · 另有 1 个档位 | 1,232 | ±3 | 48,207 | — |
| 8 | — | qwen-max-0919开源Alibaba | 1,231 | ±5 | 17,436 | — |
| 9 | — | athene-70b-0725开源NexusFlow | 1,223 | ±4 | 20,618 | — |
| 10 | — | qwen2.5-72b-instruct开源Alibaba | 1,223 | ±4 | 39,984 | — |
| 11 | — | llama-3.1-nemotron-70b-instruct开源NVIDIA | 1,217 | ±6 | 7,596 | — |
| 12 | — | llama-3.1-70b-instruct开源Meta | 1,214 | ±3 | 58,786 | — |
| 13 | — | llama-3.1-tulu-3-70b开源Ai2 | 1,210 | ±10 | 3,026 | — |
| 14 | — | gemma-2-27b-it开源Google | 1,208 | ±3 | 72,425 | — |
| 15 | — | jamba-1.5-large开源AI21 Labs | 1,205 | ±7 | 9,125 | — |
| 16 | — | llama-3.1-nemotron-51b-instruct开源NVIDIA | 1,204 | ±9 | 3,897 | — |
| 17 | — | gemma-2-9b-it-simpo开源Princeton | 1,200 | ±5 | 10,557 | — |
| 18 | — | command-r-plus-08-2024开源Cohere | 1,197 | ±6 | 10,546 | — |
| 19 | — | nemotron-4-340b-instruct开源NVIDIA | 1,196 | ±4 | 20,609 | — |
| 20 | — | llama-3-70b-instruct开源Meta | 1,195 | ±3 | 163,797 | — |
| 21 | — | qwen2.5-coder-32b-instruct开源Alibaba | 1,189 | ±8 | 5,731 | — |
| 22 | — | c4ai-aya-expanse-32b开源Cohere | 1,185 | ±4 | 26,994 | — |
| 23 | — | command-r-plus开源Cohere | 1,183 | ±3 | 80,868 | — |
| 24 | 新 | phi-4开源Microsoft | 1,182 | ±9 | 2,485 | — |
| 25 | ▼ 1 | deepseek-coder-v2开源DeepSeek | 1,182 | ±5 | 15,752 | — |
| 26 | ▼ 1 | qwen2-72b-instruct开源Alibaba | 1,181 | ±4 | 38,889 | — |
| 27 | ▼ 1 | gemma-2-9b-it开源Google | 1,179 | ±3 | 50,239 | — |
| 28 | ▼ 1 | command-r-08-2024开源Cohere | 1,176 | ±7 | 10,846 | — |
| 29 | ▼ 1 | jamba-1.5-mini开源AI21 Labs | 1,160 | ±5 | 9,269 | — |
| 30 | ▼ 1 | ministral-8b-2410开源Mistral | 1,160 | ±7 | 5,109 | — |
| 31 | ▼ 1 | qwen1.5-110b-chat开源Alibaba | 1,156 | ±3 | 27,467 | — |
| 32 | ▼ 1 | c4ai-aya-expanse-8b开源Cohere | 1,152 | ±6 | 8,797 | — |
| 33 | — | mixtral-8x22b-instruct-v0.1开源Mistral | 1,149 | ±3 | 53,788 | — |
| 34 | — | qwen1.5-72b-chat开源Alibaba | 1,148 | ±3 | 40,638 | — |
| 35 | — | command-r开源Cohere | 1,148 | ±3 | 56,380 | — |
| 36 | ▼ 4 | llama-3.1-tulu-3-8b开源Ai2 | 1,147 | ±10 | 3,077 | — |
| 37 | ▼ 1 | llama-3-8b-instruct开源Meta | 1,143 | ±3 | 109,211 | — |
| 38 | ▼ 1 | yi-1.5-34b-chat开源01 AI | 1,133 | ±4 | 25,126 | — |
| 39 | ▼ 1 | llama-3.1-8b-instruct开源Meta | 1,133 | ±3 | 52,646 | — |
| 40 | 新 | granite-3.1-8b-instruct开源IBM | 1,131 | ±11 | 2,468 | — |
| 41 | ▼ 2 | zephyr-orpo-141b-A35b-v0.1开源HuggingFace | 1,131 | ±8 | 4,860 | — |
| 42 | ▼ 2 | qwen1.5-32b-chat开源Alibaba | 1,126 | ±4 | 22,772 | — |
| 43 | ▼ 2 | phi-3-medium-4k-instruct开源Microsoft | 1,118 | ±4 | 26,112 | — |
| 44 | ▼ 1 | gemma-2-2b-it开源Google | 1,116 | ±3 | 41,868 | — |
| 45 | ▼ 3 | internlm2_5-20b-chat开源InternLM | 1,116 | ±5 | 10,599 | — |
| 46 | ▼ 2 | mixtral-8x7b-instruct-v0.1开源Mistral | 1,114 | ±0 | 76,138 | — |
| 47 | ▼ 2 | dbrx-instruct-preview开源Databricks | 1,113 | ±3 | 33,737 | — |
| 48 | ▼ 2 | granite-3.0-8b-instruct开源IBM | 1,109 | ±7 | 7,005 | — |
| 49 | ▼ 2 | qwen1.5-14b-chat开源Alibaba | 1,107 | ±4 | 18,678 | — |
| 50 | ▼ 2 | wizardlm-70b开源Microsoft | 1,105 | ±7 | 8,380 | — |
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