ModelScope APImodelscope.cn ↗
Access top trending AI models and paginated dataset listings from ModelScope.cn, including download counts, publisher info, licenses, and task types.
What is the ModelScope API?
The ModelScope.cn API exposes 2 endpoints for retrieving AI model and dataset data from China's largest open-source model community. The get_top_ten_models endpoint returns the current top ten trending models with publisher details, release dates, download metrics, and associated ML tasks. The list_datasets endpoint supports pagination to browse available datasets with license and download count data.
No input parameters required.
curl -X GET 'https://api.parse.bot/scraper/060da0cc-cb57-4a06-bd6e-6f5d2b9f5e3a/get_top_ten_models' \ -H 'X-API-Key: $PARSE_API_KEY'
Typed, relational, agent-ready
A generated client with real types, enums, and the links between objects — the structure a flat JSON response can't carry. Autocompletes in your editor and reads cleanly to coding agents.
- Fully typed · autocompletes
- Objects link to objects
- Typed errors & pagination
Typed Python client. Set up the SDK in your uv project, then pull this API’s typed client:
uv add parse-sdk uv run parse init uv run parse add --marketplace modelscope-cn-api
uv run parse add --marketplace pulls a pinned snapshot of this canonical API — it won’t change underneath you. To customize it, subscribe and swap to your own copy.
"""Walkthrough: ModelScope SDK — discover trending AI models and browse datasets."""
from parse_apis.modelscope_api import ModelScope, UpstreamError
client = ModelScope()
# Fetch trending models — single-page, capped with limit=
for model in client.models.trending(limit=5):
print(model.model_name, model.publisher.full_name, model.metrics.downloads)
# Drill into one trending model's details
top_model = client.models.trending(limit=1).first()
if top_model:
print(top_model.model_name, top_model.release_date, top_model.tasks)
# Browse datasets with pagination — limit= caps total items fetched
for ds in client.datasets.list(page_size=5, limit=3):
print(ds.name, ds.publisher, ds.license, ds.downloads)
# Typed error handling around a call
try:
results = client.datasets.list(limit=1).first()
if results:
print(results.name, results.downloads)
except UpstreamError as exc:
print(f"upstream issue: {exc}")
print("exercised: models.trending / datasets.list / typed fields / error handling")
Retrieve the top ten trending models on ModelScope.cn. Returns model name, publisher info, release date, engagement metrics, and associated tasks. The list reflects a server-side trending ranking that updates frequently; no input parameters are needed.
No input parameters required.
{
"type": "object",
"fields": {
"models": "array of model objects each containing model_name, publisher (object with name and full_name), release_date, metrics (object with downloads and likes), and tasks (array of task name strings)"
},
"sample": {
"data": {
"models": [
{
"tasks": [
"text-generation"
],
"metrics": {
"likes": 249,
"downloads": 541607
},
"publisher": {
"name": "deepseek-ai",
"full_name": "DeepSeek"
},
"model_name": "deepseek-ai/DeepSeek-V4-Flash",
"release_date": "2026.06.08"
}
]
},
"status": "success"
}
}About the ModelScope API
Trending Models
The get_top_ten_models endpoint returns an array of up to ten model objects ranked by current popularity on ModelScope.cn. Each object includes model_name, a publisher object with both name and full_name fields, a release_date, and a metrics object that surfaces download counts. Models are also tagged with their associated ML tasks, which is useful for filtering results programmatically on the consumer side.
Dataset Listings
The list_datasets endpoint returns a paginated list of datasets hosted on ModelScope. Each dataset entry includes a name field formatted as namespace/dataset, a publisher, a license string, and a downloads integer. Pagination is controlled via the page and page_size query parameters, both optional integers, allowing you to iterate through the full catalog or fetch a targeted slice.
Data Coverage
Both endpoints reflect data from ModelScope.cn, which hosts a large volume of Chinese-language and multilingual AI models and datasets. The trending model list is a snapshot of current popularity rather than a historical ranking. Dataset records expose license type directly, which is useful when assessing permissive vs. restricted use before pulling training data.
The ModelScope API is a managed, monitored endpoint for modelscope.cn — not a raw scraper you maintain. Every endpoint is automatically health-checked on a schedule, and when modelscope.cn changes and a check fails, the API is automatically queued for repair and re-verified. It is built to keep working as the site underneath it changes.
This isn't an official modelscope.cn API — it's an independent, maintained REST wrapper over public data. Where the source has no official API (or only a limited one), Parse gives you a stable contract over a source that never promised one, and keeps it current. Need a new endpoint or field? You can revise it yourself in plain English and the agent rebuilds it against the live site in minutes — contributing the change back to the shared API is free.
Will this API break when the source site changes?+
Is this an official API from the source site?+
Can I fix or extend this API myself if I need a new endpoint or field?+
What happens if I call an endpoint that has an issue?+
- Monitor which AI models are gaining traction on ModelScope to inform research or competitive analysis.
- Build a dataset discovery tool that filters ModelScope listings by license type before ingestion.
- Track publisher activity by aggregating model and dataset entries from specific namespace/publisher combinations.
- Compare download counts across top-trending models to identify adoption patterns in the Chinese AI ecosystem.
- Populate an internal model registry by syncing the top ten trending models with their task tags and release dates.
- Generate reports on open-source dataset availability, segmented by license, using paginated dataset records.
| Tier | Price | Credits/month | Rate limit |
|---|---|---|---|
| Free | $0/mo | 100 | 5 req/min |
| Hobby | $30/mo | 1,000 | 20 req/min |
| Developer | $100/mo | 5,000 | 100 req/min |
One credit = one API call regardless of which marketplace API you call. Exceeding the rate limit returns a 429 response. Authenticate with the X-API-Key header.
Does ModelScope.cn have an official developer API?+
What does the `get_top_ten_models` endpoint actually return per model?+
model_name, a publisher object with name and full_name, a release_date, a metrics object containing download counts, and an array of associated ML tasks. The list represents the current top ten trending models at query time, not a historical top list.Does the `list_datasets` endpoint support filtering by license or task type?+
name, publisher, license, and downloads per dataset and accepts only page and page_size for pagination — there are no server-side filter parameters. You can fork this API on Parse and revise it to add filtering by license or task type as an additional endpoint.Is model-level detail — such as model card content, tags, or file listings — available through this API?+
How fresh is the trending model data?+
get_top_ten_models endpoint reflects the trending state at the time of the request. ModelScope updates its trending rankings continuously, so results may shift between calls made minutes apart. There is no timestamp field on the response indicating when the ranking was last computed.