Decorative NAV chart background. Data comes from public NAV series (as of 2026-11-13). Each cell is scaled independently; heights do not represent a cross-fund performance comparison. Line colour follows the one-month direction: red means up, grey means down.
Fund Data APIThe fund data layer for AI agents
No signup, no approval queue — your key is issued on purchase
Stop scraping quarterly reports and reconciling conventions. NAV, quarterly holdings, fund managers, ratings and portfolio look-through — one API key returns structured JSON. Wire it into MCP and your agent asks directly.
- Endpoints
- 19
- Trial
- ¥9.9 / 7 days
- Concurrency
- 8
MCP · api.fundapi.cn/mcp{
"mcpServers": {
"fundapi": {
"type": "http",
"url": "https://api.fundapi.cn/mcp",
"headers": { "Authorization": "Bearer cf-YOUR_KEY" }
}
}
}Or call HTTP directly (free endpoint, no key needed)
GET /v1/fund-profile?code=110022{
"ok": true,
"endpoint": "fund-profile",
"tier": "free",
"elapsed_ms": 961,
"data": {
"code": "110022",
"name": "易方达消费行业股票",
"rate": {
"original": "1.50",
"current": "0.15",
"min_buy": "10"
}
}
}Restart your client and just ask — the agent calls these 19 tools for real data instead of guessing from memory.
Working in 3 minutes, no signup
Three endpoints return data without any key. Confirm it is what you need before you buy.
curl -s "https://api.fundapi.cn/v1/fund-profile?code=110022" -H "User-Agent: your-app"
Show the real response
{
"ok": true,
"endpoint": "fund-profile",
"tier": "free",
"elapsed_ms": 961,
"data": {
"code": "110022",
"name": "易方达消费行业股票",
"rate": {
"original": "1.50",
"current": "0.15",
"min_buy": "10"
}
}
}- Trial ¥9.9 / 7 days: search, profile, NAV, ranking, notices (6 endpoints)
- Pro ¥59/mo: + quarterly holdings, managers, estimates, dividends, fees, ratings (12 endpoints)
- Institutional ¥299/mo: + overlap, look-through, diagnostics, comparison, screener (19 endpoints)
One fund is clear. A basket is not.
You bought three funds. Are you actually holding the same stocks, how much is duplicated, what do the fees add up to? Only look-through of quarterly holdings answers that. We made it two endpoints: pass a basket of codes, get combined holdings, overlap and totals — fixed fields, conventions documented.
| Common holding | 110022 | 009265 |
|---|---|---|
| Kweichow Moutai 600519 | 9.96% | 9.82% |
| Midea Group 000333 | 9.48% | 5.90% |
| BYD 002594 | 8.64% | 4.26% |
| Shanxi Xinghuacun Fen Wine 600809 | 8.62% | 4.58% |
| CFMOTO 603129 | 6.09% | 7.92% |
| Sailun Tire 601058 | 5.61% | 6.71% |
The same data renders a full portfolio checkup
The sample below uses 3 funds at 40/30/30 weights. Every number comes from a live response (2026-11-13). With the same fields you can build charts, reports or an app — without computing look-through and overlap yourself.
- Kweichow Moutai8.84%in 3 funds
- Shanxi Fen Wine5.69%in 3 funds
- Midea Group5.56%in 2 funds
- CFMOTO4.81%in 2 funds
- BYD4.73%in 2 funds
Data captured 2026-11-13, reporting period Q3 2026. Look-through and overlap are our computed results; the response carries method and the reporting period so every figure is traceable.
Every field follows the same rules
Data is only useful if you can compute on it directly. We apply one set of field conventions to every fund:
Same name, same type, across funds
One meaning always maps to one field name and type. Change the fund or the reporting period and your code still runs.
No zero-filling, no guessing
Missing values come back as null or an empty string, with the meaning documented. An unrated fund is never turned into 0 stars.
Every number has an origin and a timestamp
Reporting period, method and data age travel with the response, so your users can audit any figure themselves.
What it can answer
Each endpoint maps to one concrete question. Not sure which one to call? Start from the question.
NAV history
TrialDaily unit NAV, accumulated NAV, daily growth rate, subscription status
/v1/fund-navQuarterly holdings
ProTop 10 holdings, weight in NAV, shares held
/v1/fund-holdingsFund managers
ProCurrent managers, tenure, AUM, capability scores, tenure return vs peers
/v1/fund-managersFund profile
FreeFees, returns for 10 periods with peer rank, AUM, asset allocation, holder structure
/v1/fund-profileIntraday estimate
ProEstimated move today — realtime and holdings-model estimates with per-stock contribution
/v1/fund-estimateDividends & splits
ProHow many payouts, amount each time, share split details
/v1/fund-dividendsThe data is public. The API is what we built.
Behind one runnable request are three things somebody had to do:
Collection and parsing are done
Quarterly holdings, NAV series, manager changes, ratings — every source is normalised into fixed fields. Missing columns follow one convention, so your code only ever sees one shape.
Conventions already decided
What unit NAV vs accumulated NAV means, which 10 periods we return, what a null rating really says — every field is pinned down in the docs so you never have to reverse-engineer it.
Timestamps travel with the data
NAV updates on T+1 evenings, holdings are dated by reporting period, estimates are labelled separately from actual NAV. The response tells you when the data is from.
Endpoints
All 19 endpoints. Required tier on the right.
| Endpoint | Name | Answers | Tier |
|---|---|---|---|
/v1/fund-search | Fund search | Find a fund by code, name or pinyin initials | Free |
/v1/fund-profile | Fund profile | Fees, returns for 10 periods with peer rank, AUM, asset allocation, holder structure | Free |
/v1/fund-screener-options | Screener options | Which types, companies, themes and ranges you can filter on | Free |
/v1/fund-nav | NAV history | Daily unit NAV, accumulated NAV, daily growth rate, subscription status | Trial |
/v1/fund-rank | Market ranking | Rank all funds by period return and see where one stands | Trial |
/v1/fund-notices | Fund notices | Index of dividend, periodic report and personnel notices | Trial |
/v1/fund-holdings | Quarterly holdings | Top 10 holdings, weight in NAV, shares held | Pro |
/v1/fund-managers | Fund managers | Current managers, tenure, AUM, capability scores, tenure return vs peers | Pro |
/v1/fund-estimate | Intraday estimate | Estimated move today — realtime and holdings-model estimates with per-stock contribution | Pro |
/v1/fund-dividends | Dividends & splits | How many payouts, amount each time, share split details | Pro |
/v1/fund-trade-rule | Trading rules | Subscription status, T+N settlement, tiered fees, channel discounts | Pro |
/v1/fund-rating | Agency ratings | Star ratings by reporting period, so you can see upgrades and downgrades | Pro |
/v1/fund-overlap | Holdings overlap | How much two or more funds share in top holdings | Institutional |
/v1/fund-portfolio-xray | Portfolio look-through | What a basket of funds actually holds once holdings are combined | Institutional |
/v1/fund-portfolio-diagnose | Portfolio diagnostics | Industry exposure, look-through top holdings, position trend, weighted fees | Institutional |
/v1/fund-holding-changes | Holdings changes | New entries, dropouts, adds and cuts this quarter | Institutional |
/v1/fund-manager-history | Manager history | How often the manager changed and what each tenure returned | Institutional |
/v1/fund-compare | Fund comparison | Side-by-side metrics for several funds, plus pairwise overlap | Institutional |
/v1/fund-screener | Market screener | Filter the whole market by type, company, theme, return range and rating | Institutional |
Pricing
¥9.9 for 7 days is enough to judge the data. Paid tiers are subscriptions, not one-off purchases.
Trial
One-off, not stackable
7 days to check whether the data is accurate and complete
Try 7 daysGet it working first, then decide
Search, profile and screener options work without a key. To see NAV and rankings, ¥9.9 for 7 days is enough to make up your mind.