DeepSeek R1 Distill Qwen 32B
DeepSeekDeepSeek开源权重MIT · 商用许可
描述
DeepSeek-R1 is the first-generation reasoning model built atop DeepSeek-V3 (671B total parameters, 37B activated per token). It incorporates large-scale reinforcement learning (RL) to enhance its chain-of-thought and reasoning capabilities, delivering strong performance in math, code, and multi-step reasoning tasks.
发布日期
2025-01-20
参数规模
32.8B
上下文长度
—
支持模态
text
能力雷达图
27
general
27
coding
69
reasoning
44
science
56
agents
0
multimodal
排行榜排名
基准测试分数 (LLM Stats)
(LLM Stats (zeroeval))Math
MATH-500
94.3%自报
AIME 2024
83.3%自报
Reasoning
GPQANYU + Cohere + Anthropic (2023)
62.1%自报
LiveCodeBench
57.2%自报
AA 评测指数
(Artificial Analysis)Math 500(OpenAI (2024), subset of Hendrycks et al. MATH (2021))94.1
Mmlu Pro(TIGER-Lab (Univ. of Waterloo, Toronto, CMU, 2024))73.9
Aime(MAA (Mathematical Association of America))68.7
Math Index(Artificial Analysis)63.0
Aime 25(MAA (Mathematical Association of America))63.0
Gpqa(NYU + Cohere + Anthropic (2023))61.5
Livecodebench(UC Berkeley + MIT + Cornell (2024))27.0
Ifbench(Google Research (2023))22.9
Lcr(Artificial Analysis)8.7
Intelligence Index(Artificial Analysis)8.4
Hle(Center for AI Safety + Scale AI (2025))4.6
LLM Stats 分类评分
(LLM Stats (zeroeval))Math90
Reasoning70
General70
Physics60
Biology60
Chemistry60
Code60
定价
输入价格免费
输出价格免费
混合价格(3:1)免费
速度
Tokens/秒0.0
首Token延迟0.00s
首回答延迟0.00s
供应商价格排行
供应商价格排行
1 个供应商
供应商输入输出
1Alibaba (China)
$0.287
$0.861
比较该模型在不同 API 供应商之间的定价。