Modelers 魔乐社区 (modelers.cn, Huawei/Ascend ecosystem)
- Free allowance
- Not published. The community is an explicit Hugging-Face-shaped hub — models, datasets, and Spaces, plus its own client (openMind Hub Client) and library (openMind) targeting Ascend NPUs.
- Verdict
- caveatusable, with a named limit or an unprobed read path
- Cost
- free tier
- Identity
- emailno card, phone required
- Binding limit
- per-object-capNo per-file cap, account quota, retention, or rate limit found
- Evidence
- inferred3 sources; reasoned from other evidence
- Freshness
- stale on 2026-11-21checked 2026-10-21 (inherited); stale on 2026-11-21
| Free tier | Not published. The community is an explicit Hugging-Face-shaped hub — models, datasets, and Spaces, plus its own client (openMind Hub Client) and library (openMind) targeting Ascend NPUs. Its contribution guide instructs 'git lfs track' for '100MB以上' files, which confirms a real LFS backend rather than a web-only uploader. unit GB; metered on unstated |
|---|---|
| Price | none stated |
| Limits |
Unverified. git-LFS 100 MB+ tracking threshold documented. No per-file cap, account quota, retention, or rate limit found. Upload requires an access token with Write scope created in the personal centre. |
| Identity | email email + +86 phone (Huawei-adjacent Chinese community); access token required for any write. |
| Egress | No egress meter documented; models download through both the web and the openMind client anonymously. CN-network served. |
| Per-object cap | No per-file cap, account quota, retention, or rate limit found |
| Retention | not stated |
| Interface | git part of the text of Free tier, above |
| Tooling | First-party openMind Hub Client (pip) with upload/download by repo_id, plus git+LFS, plus a web UI. Documented at modelers.cn/docs/zh/openmind-hub-client. This is a genuine HF drop-in: repo types are 'model', 'dataset', 'space', same as HF. |
| Composes with | N repos per account, free. The strategic value is that it is the domestic Ascend/MindSpore hub, so it is the natural place a Chinese-ecosystem user would park multi-hundred-GB checkpoints. |
| Reliability | API surface is opaque: https://modelers.cn/api/v1/models returns 404, and model detail pages return a 1,638-byte SPA shell regardless of path — so anonymous programmatic access is effectively absent; you need the openMind client. |
| Parsed numbers |
|
Evidence 3
| date | source | what it proved |
|---|---|---|
| 2026-10-21 | modelers.cn/docs/zh/best-practices/community_contribution/mo | read live: '在本地Git仓库中追踪大型文件（100MB以上）', git lfs upload flow, write-scoped access token required |
| 2026-10-21 | modelers.cn/ | read live: 'Models', 'Spaces', 'openMind Library', 'openMind Hub Client — Easily search for, upload, and download pre-trained models'; dataset upload guide dated 2024-09-22 |
| 2026-10-21 | modelers.cn/api/v1/models?per=2 | probe: GET -> 404; GET https://modelers.cn/models/OpenMind/DeepSeek-R1-Distill-Qwen-7B -> 200 but 1,638-byte SPA shell; POST .../info/lfs/objects/batch -> 404 (returns the SPA shell, not LFS JSON) |
Unknown values 3
- per object cap notes: the field may be stated in a source that was not read
- retention notes: the field may be stated in a source that was not read
- retention stated: the field may be stated in a source that was not read