本地商家发现
Use Google Maps endpoints to turn 本地商家发现 into repeatable API requests with documented inputs and JSON responses.
从 Google Maps 采集本地商家和地点数据，用于位置情报工作流。
通过文档化的 Crawlora API 端点获取 Google Maps 的地点详情和本地搜索结果。
Endpoint families
2
Documented params
10
Examples
4
Live catalog snapshot
Active endpoints
4
Methods
GET, POST
Required params
8
Schema refs
4
{
"platform": "Google Maps",
"endpoint": "google-map-place",
"method": "GET",
"path": "/google/map/place/{place_id}",
"auth": "apiKey"
}Related use cases
Free tools
Use cases
通过文档化的 Crawlora API 端点获取 Google Maps 的地点详情和本地搜索结果。
Use Google Maps endpoints to turn 本地商家发现 into repeatable API requests with documented inputs and JSON responses.
Use Google Maps endpoints to turn 地点详情数据增强 into repeatable API requests with documented inputs and JSON responses.
Use Google Maps endpoints to turn 位置情报 into repeatable API requests with documented inputs and JSON responses.
Managed execution
Every figure below is read from the live Google Maps endpoint catalog — 4 endpoints, 10 documented request parameters, and 4 published response schemas — the same catalog Docs and Playground run against.
4 documented Google Maps endpoints, grouped into 2 request families — Place and Search.
10 request parameters are documented across those Google Maps endpoints, 8 of them required — the full input contract is public before you write any integration code.
4 of the 4 Google Maps endpoints ship a recorded example response, and 4 carry a documented response schema — you can code against the real JSON before the first request.
Google Maps endpoints document their error responses (400, 429 and 500) alongside the success schema, so a block, a rate limit, or a missing record comes back as a typed error rather than silently empty data.
4 hosted MCP tools back the Google Maps endpoints, so an agent can call the same routes with the same parameters and the same JSON contract, with no custom glue.
Coverage map
These cards are generated from the active endpoint catalog, so the landing page reflects the same API surface used by Docs and Playground.
/google/map/place/{place_id}
/google/map/search
Endpoint catalog
/google/map/place/{place_id}Returns detailed information for a specified place_id. Rate limit is enforced at 1 request per second.
Response notes
- `rating` is the aggregate Google rating (`1.0`–`5.0`), or `null` when the place has no aggregate rating on Google (too few reviews, or a place type Google does not rate) — never a literal `0`. Read it together with `review_count`.
MCP tool google_map_place
/google/map/place/{place_id}/photosReturns the photos Google publishes for a specified place_id — the imagery shown on the place's Google Maps page, typically dozens of images for a well-covered business. Each entry carries the image URL as served plus its pixel dimensions when reported; swap the trailing size suffix on the URL (e.g. `=w203-h100-k-no`) to request other dimensions. Contributor avatars and review-attached photos are excluded. This is the place page's image set, not a paginated archive feed. Rate limit is enforced at 1 request per second.
Response notes
- `url` is Google's image URL exactly as served, including its trailing size suffix (for example `=w203-h100-k-no`). Swap that suffix to request other dimensions. - `width` and `height` are the pixel dimensions of that URL when Google reports them. `0` means the dimension was not captured, not that the image is empty. - Contributor avatars and review-attached photos are excluded; this endpoint returns place imagery only. Photos attached to individual reviews belong to the review that carries them. - Each distinct image appears exactly once, at the largest size Google offers for it. Google's payload carries the same photo at several sizes; those variants are collapsed rather than returned as separate entries. - This is the place page's image set, **not** a paginated archive feed. There is no cursor to walk, so the response is the full set this endpoint can return for the place. - A place with no published photos returns an empty `data` array, not an error. Example response: ```json { "code": 200, "msg": "ok", "data": [ { "url": "https://lh3.googleusercontent.com/gps-cs-s/AHRPTWlbepbs69mWZVN7SNi4FYseIr1JvRCpGHgpFfN5wOf02mKO=w203-h100-k-no", "width": 203, "height": 100 }, { "url": "https://lh3.googleusercontent.com/gps-cs-s/AHRPTWntpDTSCFVQdK1TmGSIhUaUWhoH45c1klBfg6se1L-m4xSO=w408-h306-k-no", "width": 408, "height": 306 } ] } ```
MCP tool google_map_place_photos
/google/map/place/{place_id}/reviewsReturns the reviews Google shows on a specified place_id's Google Maps page — typically the 8 most relevant, each with its rating, text, reviewer, timestamp, and any photos the reviewer attached. Photo-only reviews return an empty `text`. This is the place page's first page of reviews, not the full review archive. Rate limit is enforced at 1 request per second.
Response notes
- This is the place page's **first page** of reviews, not the full review archive. There is no cursor to page through, so the response is the complete set this endpoint can return for the place. Use `review_count` on `/google/map/place/{place_id}` for the place's total review tally. - `text` is empty for a photo-only review — Google allows a review with photos and a rating but no written body. That is not a parse failure. - `published_at` is RFC3339 UTC. It is empty when Google did not report a timestamp. - `relative_time` is Google's own wording ("6 months ago") in the requested language, so it is a display string rather than something to parse — use `published_at` for date logic. - `photos` are the images attached to that specific review. They never overlap with `/google/map/place/{place_id}/photos`, which returns the place's own gallery; a review's photos belong to the review. - `author_url` is normalized to drop Google's `?hl=` suffix, so the same reviewer yields the same URL regardless of the request language. - `author_review_count` is how many reviews that contributor has written overall. `0` means Google did not report it, which is normal for a syndicated review. - `source` names the site the review came from. Google's place pages syndicate reviews from partner sites — common on hotels — so this is `Google` for a native Google review and the partner's name (for example `Tripadvisor`, `TrustYou`) otherwise. Syndicated reviews still carry a rating, text, author name, and timestamp, but have no Google contributor id, so `author_id` is empty and `author_url` points at the partner site rather than a Google contributor page. - A place nobody has reviewed returns an empty `data` array, not an error. Example response: ```json { "code": 200, "msg": "ok", "data": [ { "id": "Ci9DQUlRQUNvZENodHljRjlvT201dFRHSk1aRU54T0U5cmRFVjNlbGRmU1RJNFkzYxAB", "rating": 5, "text": "I never knew pork could taste this good. We tried the special loin and it was worth the queue.", "relative_time": "6 months ago", "published_at": "2026-01-01T14:39:25Z", "author_name": "Supisara Darnchaloemwong", "author_id": "112997636113922766481", "author_url": "https://www.google.com/maps/contrib/112997636113922766481", "author_photo": "https://lh3.googleusercontent.com/a-/ALV-UjWc6XFX9IKHJiQ4vZoY7o_EVCYkjMbWjGfdPZrLEw", "author_review_count": 23, "source": "Google", "photos": [ { "url": "https://lh3.googleusercontent.com/grass-cs/ACvplmM2ICW60UtuEQbkJLN-ot5vq3pCknhfn=w1207-h1262-k-no", "width": 1207, "height": 1262 } ] } ] } ```
MCP tool google_map_place_reviews
/google/map/searchReturns results from Google Maps based on search options. Rate limit is enforced at 1 request per second.
Response notes
- `rating` is the aggregate Google rating (`1.0`–`5.0`), or `null` when the place has no aggregate rating on Google (too few reviews, or a place type Google does not rate) — never a literal `0`. Read it together with `review_count`.
MCP tool google_map_search
Related APIs
Maps & Local Data
Search ready-made scraped datasets for local business research with filters, facets, geo queries, and repeatable API output.
Maps & Local Data
Normalize location search, coordinate lookup, and OSM identifier enrichment through API-key protected geocoding endpoints.
Maps & Local Data
Compare cities and countries with normalized Numbeo cost-of-living and quality indices. Retrieve city or country detail and current or historical rankings for location research, relocation products, market analysis, and local-data enrichment.
How to scrape Google Maps
Google Maps 拥有最丰富的公开本地商家数据集，但它是为人类浏览渲染的，而非为 API 设计的。Crawlora 的 Google Maps 端点将地点、评分、评论和联系方式返回为结构化 JSON。
使用查询词加地点进行搜索（例如“奥斯汀的咖啡店”），或按名称或 ID 查找单个地点。
将查询词和地点（或地点标识）传给 Crawlora 的 Google Maps 端点——无需维护无头浏览器。
为每条结果获取商家名称、地址、电话、评分、评论数量、营业时间、坐标和网站。
按域名去重，用 Contact API 为每个网站补充邮箱和社交账号信息，或定期重新运行以跟踪新增和变更的商家信息。
FAQ
将查询词加地点（或某个具体地点）传给 Crawlora 的 Google Maps 端点。你会得到结构化 JSON——商家名称、地址、电话、评分、评论、营业时间、坐标和网站——无需维护浏览器自动化。请采集公开数据并遵守 Google 的条款。
可以——Crawlora 的 Google Maps 端点作为一个 Google Maps 数据 API，提供地点、评分、评论、营业时间、坐标和本地商家数据，均为结构化 JSON。它返回 Maps 的结果，且无需管理官方 Places API 配额或无头浏览器。
Google Maps 只列出每个商家的网站和电话，不包含邮箱。将 Maps 的结果与 Crawlora 的 Contact API 结合，即可从每个商家的网站提取邮箱和社交账号——这是外联营销中缺失的最后一环。
在遵守 Google 使用条款、robots 指令、速率限制和适用法律的前提下，抓取公开可见的地点和商家数据通常是被允许的。这适用于你有权采集的公开数据。