OpenXLab 浦源
openxlab.org.cn · Storage · 1 record · JSON
OpenXLab 浦源 (Shanghai AI Lab) — model + dataset centers Storage
caveatemailno cardphoneI
| Free tier | Not published, and this is the honest finding. The platform's own intro describes the model center as building an object-storage layer specifically for large files — verbatim: '基于平台内的命令行工具，便于 AI 开发者上传和发布模型文件，搭建对象存储体系，提供大文件存储能力，快速上传下载功能，便于 AI 开发者进行模型存储' (object storage system providing large-file storage capability). The dataset center supports private datasets as well as public ones. But no GB number appears on any page I could reach; the openxlab.org.cn docs tree 504'd repeatedly and the upload pages contain no quota statement. unit unlimited; metered on stored-bytes |
|---|---|
| Price | none stated |
| Per-object cap | not stated |
| Retention | not stated |
| Egress | No egress meter found; public models and datasets are anonymously downloadable. Reads are CDN-served. Unverified at the header level because model-detail pages timed out from this network. |
| Identity | email email, plus likely +86 phone — Chinese platform, and prior rounds recorded the pans in this family demanding real-name verification. |
| Binding | none no constraint in the record binds at this scale; the record carries no limit of a kind the vocabulary names |
| Interface | rest, git, cli rest |
| Tooling | pip install openxlab, then first-party subcommands: openxlab dataset upload-file, openxlab dataset upload-folder, openxlab model upload ... with --dataset-repo/--source-path/--target-path. Python SDK mirrors the CLI (upload_file / upload_folder). Also git+LFS. Real, named, documented client — this passes the tooling filter cleanly. |
| Composes with | Repos are per-user and per-org and free to create, so capacity is N repos. OpenXLab is the Shanghai AI Lab's public front door (InternLM, OpenDataLab datasets), so its per-repo sizes are already in the tens-of-GB range. Multiplying across accounts is easy but the per-account ceiling is unknown. |
| Reliability | Model pages return 302 to login for detail views. |
| Notes | Not published, and this is the honest finding. The platform's own intro describes the model center as building an object-storage layer specifically for large files — verbatim: '基于平台内的命令行工具，便于 AI 开发者上传和发布模型文件，搭建对象存储体系，提供大文件存储能力，快速上传下载功能，便于 AI 开发者进行模型存储' (object storage system providing large-file storage capability). The dataset center supports private datasets as well as public ones. But no GB number appears on any page I could reach; the openxlab.org.cn docs tree 504'd repeatedly and the upload pages contain no quota statement. |
| Evidence rung | inferred One page read moves it to documented. |
| Checked | 2026-10-21 (inherited); stale on 2026-11-20 |
Evidence (3)
| date | source | what it proved |
|---|---|---|
| 2026-10-21 | openxlab.org.cn/docs/intro.html | read live: model center provides '对象存储体系，提供大文件存储能力' and the dataset center supports '私有数据集的上传、管理和发布' |
| 2026-10-21 | opendatalab.com/docs/datasets/上传数据集.html | read live: three upload modes (file / folder / archive with auto-extract), plus CLI and Python SDK (pip install openxlab, openxlab dataset upload-folder) |
| 2026-10-21 | withheld | unverified: no quota page exists that I could reach; openxlab.org.cn docs paths 504'd repeatedly and opendatalab.com API paths timed out at 25-30 s |
Unknown or withheld values (6)
$.evidence[2].urlprivate-url$.free_tier.notesprocess$.per_object_cap.notesnot-extracted$.per_object_cap.statednot-extracted$.retention.notesnot-extracted$.retention.statednot-extracted