Export microCMS to CSV or JSON: Fetch All Content and Move Images

To export microCMS content, use the dashboard’s CSV export on Team or higher, or fetch every item through the Content API with the JavaScript SDK’s getAllContents method. Hobby users can retrieve JSON through the API and convert it to CSV themselves. Save API schemas and actual image or file downloads separately from content records.

This article is written by the genko.me team. We checked the official documentation on October 23, 2026. microCMS’s built-in CSV workflow is useful for teams that want spreadsheet exports without maintaining a script; exporting data does not require switching CMS products.

Can microCMS export CSV from the dashboard?

Official CSV export documentation lists Team, Business, and Enterprise as supported plans. Hobby does not include this dashboard feature. Content export and API schema export are separate operations.

1. Open the content list for the API you want to export.

2. Choose “All Contents” for a complete export and clear search terms and filters.

3. Select CSV Export, check the export name and scope, and start the job.

4. Download the result from the completion email. The link works only for the user who started the export and expires after 24 hours.

The active view’s conditions and sort order determine what is exported. An item that is both published and being edited appears as two rows sharing an ID: one published version and one draft version. Count unique IDs separately from rows. References or other values that the exporting user cannot read can appear as NO_READ_PERMISSION. Avoid editing during the export because changes may affect the result.

How do you fetch all microCMS content beyond 100 items?

The list endpoint is GET https://{service}.microcms.io/api/v1/{endpoint}. Its default limit is 10 and its maximum is 100. Setting limit=1000 is not a way to retrieve every record.

For manual pagination, request limit=100 with offset=0, then offset=100, then offset=200. Append each response’s contents array. Stop when the next starting position reaches or exceeds totalCount, which counts items matching the request conditions. Leave out filters and q for an unfiltered export, and avoid fields when the goal is preserving every field.

Export every item to JSON with TypeScript

The official JavaScript SDK provides getAllContents({ endpoint, queries?, customRequestInit? }). It returns an array and handles pagination. Its queries type excludes limit, offset, and ids.

The example below targets list APIs. Use getObject separately for an object API, such as a single site settings record. In a working directory using Node.js 22 or newer, install the SDK and script runner:

npm install microcms-js-sdknpm install --save-dev tsx

Save this as export.ts. Supply the three environment variables through your execution environment or its secret settings. Keep credentials out of the script and Git.

import { writeFile } from "node:fs/promises";import { createClient } from "microcms-js-sdk";
async function main() {  const serviceDomain = process.env.MICROCMS_SERVICE_DOMAIN;  const apiKey = process.env.MICROCMS_API_KEY;  const endpoint = process.env.MICROCMS_ENDPOINT;  if (!serviceDomain || !apiKey || !endpoint) {    throw new Error("Missing microCMS environment variables");  }  const client = createClient({ serviceDomain, apiKey, retry: true });  const contents = await client.getAllContents<Record<string, unknown>>({    endpoint,    queries: { depth: 0 },  });  await writeFile("export.json", JSON.stringify({ contents }, null, 2) + "\n");  console.log(`Exported ${contents.length} items`);}main().catch((error) => {  console.error("Export failed:", error instanceof Error ? error.message : error);  process.exitCode = 1;});

Run it after setting MICROCMS_SERVICE_DOMAIN to the service name, rather than the full URL, and MICROCMS_ENDPOINT to an endpoint such as blog:

npx tsx export.ts

The resulting export.json wraps the returned array in contents. With depth: 0, references are represented by IDs rather than expanded records. Export each referenced API separately so those IDs remain useful. Keeping references compact also makes later transformations easier to inspect.

Which API permissions are needed for an export?

Use a key with GET permission only for the APIs you need. If drafts or closed content are required, enable their respective bulk retrieval permissions as described in the API key documentation. Keep both the privileged key and the exported files private.

Do not treat a successful list export as a complete archive of published versions, unpublished changes, and revision history. Confirm which values actually came back, especially for a published item with pending edits. The dashboard CSV’s separate version rows are also worth checking when preserving editorial state matters.

Pause editing before a large export, then check item counts and duplicate IDs. Offset pagination across changing records can miss or repeat items. A matching count alone does not prove that every body or attachment is intact.

What if the export hits size or rate limits?

microCMS documents an approximately 5 MB maximum response size. If getAllContents fails because a page is too large, switch to getList with a smaller limit and advance offset against totalCount yourself. Preserve the full export by reducing page size rather than dropping fields that you still need.

The documented GET rate limit is 60 origin requests per second per service. Other workloads contribute to that total; responses served from the CDN edge cache are excluded from the origin count. Run export requests sequentially, and wait before retrying 429 responses or transient failures. These figures are from the official limitations page, checked in October 2026.

How to convert microCMS JSON to CSV

Save the following as to-csv.ts, next to export.json, and run npx tsx to-csv.ts. It discovers top-level columns across the entire array. Arrays and objects remain JSON strings inside cells, preserving their structure for inspection without pretending that they are ordinary spreadsheet columns.

import { readFile, writeFile } from "node:fs/promises";
async function main() {  const { contents } = JSON.parse(await readFile("export.json", "utf8")) as {    contents: Record<string, unknown>[];  };  const columns = [...new Set(contents.flatMap((row) => Object.keys(row)))];  const cell = (value: unknown): string => {    let text = value == null ? "" : typeof value === "object"      ? JSON.stringify(value) : String(value);    if (/^[\s]*[=+\-@]/.test(text)) text = "'" + text;    return `"${text.replace(/"/g, '""')}"`;  };  const rows = [columns.map(cell).join(","), ...contents.map((row) =>    columns.map((key) => cell(row[key])).join(","))];  await writeFile("export.csv", "" + rows.join("\r\n") + "\r\n", "utf8");}main().catch((error) => {  console.error("CSV conversion failed:", error instanceof Error ? error.message : error);  process.exitCode = 1;});

The converter quotes every cell, escapes quotes, preserves embedded newlines, and adds a UTF-8 BOM. It prefixes cells that could be interpreted as spreadsheet formulas with an apostrophe. This changes some values, including negative numbers, so retain JSON as the authoritative copy. This CSV is for inspection; it is not a ready-made import template for another CMS.

Do not flatten repeaters just to make a spreadsheet look cleaner. First decide whether the destination will represent them as structured blocks, separate records, or another supported field type. Keep the untouched source JSON so that transformation choices remain reversible.

How do you export microCMS images and media?

Collect each image object’s url, image arrays, and image src values embedded in rich text. This covers assets referenced by the content you exported. It will not find unused media in the library.

For a library inventory, microCMS’s media export guidance points to the Management API’s GET /api/v2/media. Enable the “Get media” permission. Download each file from its returned URL; saving the inventory alone does not back up the file contents.

The media list endpoint accepts a maximum initial limit of 100. Subsequent pages use token, rather than content-style offset pagination. The token has a documented 15-second expiry, so collect the inventory first and download binaries afterward. A token can be present on the first response even when there are no additional results: use the fetched count and totalCount, rather than token presence alone, to determine completion.

The Management API limit is 10 requests per 10 seconds per IP address. This is a different scope from Content API GET limits. Keep a mapping between original URLs and downloaded filenames, and include PDFs or other file types where used. During migration, check both dedicated image fields and old media URLs embedded in article bodies.

How to import a microCMS export into genko.me

genko.me accepts microCMS schema JSON containing apiFields. This is schema compatibility, rather than a dedicated importer that automatically converts every content value and copies media. Follow the destination format in Import and export.

1. Export the schema from microCMS’s API Settings → API Schema screen. Load it in genko.me’s schema editor and review the converted fields. Unsupported custom and repeater structures need redesigning.

2. Match field IDs and types, then construct content JSON with only the fields intended for the destination. Remove metadata such as revisedAt when it is not a destination field.

3. Upload images to genko.me and replace source image objects with storage values containing destination media IDs. Create referenced records first and map their IDs explicitly.

4. Open Import on the destination API, select the JSON file, and initially import as drafts.

5. Check counts, IDs, bodies, images, and links before publishing.

For a destination with just title and body, a transformed input looks like this:

{  "contents": [    { "id": "first-post", "title": "First post", "body": "<p>Article body.</p>" }  ]}

Keep the original export separately. As of October 2026, genko.me accepts up to 10 MB per file and 1,000 items per import. Split large exports to satisfy both limits. Eligible content IDs can be retained, but a duplicate existing ID stops an import; items already saved remain saved.

Source publication states, creation timestamps, and scheduled publication settings are not restored. If the old publication date must appear on the site, migrate it into a dedicated date field. The response value reference also explains why delivery responses and stored media values can differ.

Use our microCMS comparison for product selection and free-plan limits guide for Hobby planning. If your only requirement is a periodic export, continuing to use microCMS can be the practical choice.

FAQ

Can I export microCMS content on the free plan?

Hobby users can retrieve content through the GET API and save JSON or generate CSV themselves. Dashboard CSV export requires Team or higher. API and asset retrieval remain subject to the service’s usage conditions.

How can I get more than 100 microCMS items?

Use getAllContents, or paginate getList with limit and offset. “All” means all records readable by that API key under the request conditions, rather than every historical version in the service.

Can I export drafts and closed content?

The API requires GET plus the corresponding bulk retrieval permissions. For published items with pending edits, inspect the returned values and the dashboard CSV’s separate version rows before claiming that both versions have been preserved.

Does a CSV export back up images?

A CSV or JSON export containing image URLs does not save the image files themselves. Download the assets separately and check embedded images in rich text as well as dedicated image fields.

Sources

Checked October 23, 2026.

• microCMS: CSV export

• microCMS: List GET API

• microCMS: JavaScript SDK, implementation, and types

• microCMS: API keys and permissions

• microCMS: Media export

• microCMS: Media list API

• microCMS: Limitations

• microCMS: Schema export

• genko.me: Import and export

• genko.me: Plans and limits