API、托管或本地运行
DeepSeek-R1-Distill-Qwen-32B
Transformers / vLLM / SGLang / Docker Model Runner
接入说明：DeepSeek-R1-Distill-Qwen-32B model card ↗选择要核对的信息，直接进入对应章节；各项事实均保留版本、条件和来源。
7/14 项有公开值 · 当前版本 DeepSeek-R1-Distill-Qwen-32B · 数据截止 2026-10-18
页面只展示已审核 Claim；待审核数据不会被当作公开事实。只展示已审核事实；未公开、冲突和过期状态不会被补猜。
只判断选型资料是否足够，不评价模型能力高低。
已审核的公开资料尚不能确认 API 或本地部署路径，需要补充接入证据。
DeepSeek R1 Distill Qwen 32B 是 DeepSeek 的 DeepSeek R1 Distill 模型，年鉴当前记录 1 个版本，最新公开记录为 DeepSeek-R1-Distill-Qwen-32B。
权重为公开权重，代码为暂无已审核记录，训练数据为暂无已审核记录，部署方式为本地部署；这些维度分别判断，不能合并成单一“开源/闭源”结论。
当前公开快照的上下文窗口为 131.1K tokens，架构为 Qwen2ForCausalLM (dense Transformer)；这里只使用通过审核且来源允许公开采集的 Claim。
7 个维度暂无可公开投影的审核记录
补充来源或线索会进入投稿队列，经编辑核验后才会成为公开事实。当前来源记录 2 个。
技术事实、业务价值、企业治理和证据质量分区呈现
厂商、家族、API ID、日期与生命周期
Dense/MoE、参数、层数、专家与 Tokenizer
预训练、SFT、偏好优化与安全对齐
教师/学生、基础模型、轨迹和限制
权重、代码、数据、许可证、部署与 API
输入输出、缓存、Batch 与部署成本边界
数据、权限、审计、合规、安全与人工兜底
5 条已审核事实 · 全部来自公开来源闸门通过的记录
尚无已审核公开记录的维度：02 能力与任务、07 部署与接入、09 性能与评测、11 业务场景、12 落地与提效。查看资料缺口 ↑
关系只来自年鉴数据库中已登记的归属字段。
当前最新版本的 approved Claim 投影 · 1 条历史事实正在复核
下载、API、客户端和更新内容只展示已登记的来源事实；未登记不作推断。
Transformers / vLLM / SGLang / Docker Model Runner
接入说明：DeepSeek-R1-Distill-Qwen-32B model card ↗保留发布日期精度、生命周期节点和原始来源，未公开日期不会被推断。
模型版本 · active
显示最新核验的 5 条 · 共 5 条已审核事实
The official DeepSeek AI repository identifies the model as DeepSeek-R1-Distill-Qwen-32B. Its public configuration declares Qwen2ForCausalLM with max_position_embeddings 131072; published BF16 safetensor metadata totals 32,763,876,352 parameters. The model card provides Transformers, vLLM, SGLang and Docker Model Runner usage paths.
The official DeepSeek AI repository identifies the model as DeepSeek-R1-Distill-Qwen-32B. Its public configuration declares Qwen2ForCausalLM with max_position_embeddings 131072; published BF16 safetensor metadata totals 32,763,876,352 parameters. The model card provides Transformers, vLLM, SGLang and Docker Model Runner usage paths.
1 个版本 · 每项可回到来源与技术解释
高置信度，已绑定来源与证据。
条件：Architecture class from the official configuration.
来源：DeepSeek-R1-Distill-Qwen-32B model card ↗The official DeepSeek AI repository identifies the model as DeepSeek-R1-Distill-Qwen-32B. Its public configuration declares Qwen2ForCausalLM with max_position_embeddings 131072; published BF16 safetensor metadata totals 32,763,876,352 parameters. The model card provides Transformers, vLLM, SGLang and Docker Model Runner usage paths.
高置信度，已绑定来源与证据。
条件：Maximum model position length from the official configuration.
来源：DeepSeek-R1-Distill-Qwen-32B model card ↗The official DeepSeek AI repository identifies the model as DeepSeek-R1-Distill-Qwen-32B. Its public configuration declares Qwen2ForCausalLM with max_position_embeddings 131072; published BF16 safetensor metadata totals 32,763,876,352 parameters. The model card provides Transformers, vLLM, SGLang and Docker Model Runner usage paths.
引用文本与接口返回都只包含已审核公开事实
DeepSeek DeepSeek R1 Distill Qwen 32B｜最新版本 DeepSeek-R1-Distill-Qwen-32B｜已审核事实 5 条｜公开来源 2 个｜数据截止 2026-10-18｜AI 年鉴｜https://ainianjian.com/models/model.deepseek.deepseek-r1-distill-qwen-32b
公开接口返回该模型的版本、已审核事实 5 条及其来源信息，可直接用于脚本或对比。
https://api.ainianjian.com/api/models/model.deepseek.deepseek-r1-distill-qwen-32b ↗curl -s 'https://api.ainianjian.com/api/models/model.deepseek.deepseek-r1-distill-qwen-32b'
技术规格使用当前版本已审核且来源允许公开采集的事实；模型级开放信息与跨版本地区记录分别标注范围。请按每项适用范围核对，未登记字段显示「未公开」，不使用估算值填充。
The official DeepSeek AI repository identifies the model as DeepSeek-R1-Distill-Qwen-32B. Its public configuration declares Qwen2ForCausalLM with max_position_embeddings 131072; published BF16 safetensor metadata totals 32,763,876,352 parameters. The model card provides Transformers, vLLM, SGLang and Docker Model Runner usage paths.
The official DeepSeek AI repository identifies the model as DeepSeek-R1-Distill-Qwen-32B. Its public configuration declares Qwen2ForCausalLM with max_position_embeddings 131072; published BF16 safetensor metadata totals 32,763,876,352 parameters. The model card provides Transformers, vLLM, SGLang and Docker Model Runner usage paths.
DeepSeek official API documentation labels the DeepSeek-R1 release as 2025/01/20 and states that DeepSeek-R1 plus six dense distilled models based on Qwen2.5 and Llama3 were open-sourced.
高置信度，已绑定来源与证据。
条件：Explicit usage paths on the official model card.
来源：DeepSeek-R1-Distill-Qwen-32B model card ↗The official DeepSeek AI repository identifies the model as DeepSeek-R1-Distill-Qwen-32B. Its public configuration declares Qwen2ForCausalLM with max_position_embeddings 131072; published BF16 safetensor metadata totals 32,763,876,352 parameters. The model card provides Transformers, vLLM, SGLang and Docker Model Runner usage paths.
高置信度，已绑定来源与证据。
条件：The official R1 announcement explicitly includes this distilled model.
来源：DeepSeek-R1 Release ↗DeepSeek official API documentation labels the DeepSeek-R1 release as 2025/01/20 and states that DeepSeek-R1 plus six dense distilled models based on Qwen2.5 and Llama3 were open-sourced.
高置信度，已绑定来源与证据。
条件：Published BF16 safetensor parameter total; this is not a hardware memory estimate.
来源：DeepSeek-R1-Distill-Qwen-32B model card ↗The official DeepSeek AI repository identifies the model as DeepSeek-R1-Distill-Qwen-32B. Its public configuration declares Qwen2ForCausalLM with max_position_embeddings 131072; published BF16 safetensor metadata totals 32,763,876,352 parameters. The model card provides Transformers, vLLM, SGLang and Docker Model Runner usage paths.