Hypotenuse AI

The Webflow CMS stores and renders blog posts, but it does not write them. Teams that draft articles in a separate tool often paste each one into a rich text field by hand, and content pasted from Microsoft Word comes through corrupted, which leaves someone reformatting it before the post can go out.

Hypotenuse AI connects to Webflow directly, so a finished blog article moves into a Collection in one export. You pick the blog Collection and map the article title, content, outline and keywords to Collection fields. Bulk product copy and metadata take a different route: a CSV export from Hypotenuse AI, then a CMS import or your own code against the Webflow Data API.

Content marketers and SEO specialists who publish to a Webflow blog get the most direct benefit. Agencies can link several Webflow workspaces, or several sites in one workspace, through one connected Webflow account, and write in a separate brand voice for each client. Ecommerce teams with a developer can route bulk product descriptions into Webflow Ecommerce through the Data API, since Hypotenuse AI does not document a native Webflow product sync.

How to integrate Hypotenuse AI with Webflow

What is Hypotenuse AI? Hypotenuse AI describes itself as an AI-first PIM (product information management) platform for ecommerce. Its catalog tools enrich product data and write product descriptions in bulk. It also includes a blog writer that exports finished articles to Webflow.

Pick the route that matches the content type and how much of it you publish:

• Export to Webflow: The integration built into Hypotenuse AI sends blog articles to a Collection without any code.
• CSV export and CMS import: Bulk product descriptions and blog batches leave Hypotenuse AI as CSV files that the Webflow CMS can import into a Collection.
• DOCX or HTML paste: One-off documents go into a rich text field, and HTML press releases go into a Code Embed element.
• Webflow Data API: Your own server-side code writes exported content into Collections or Ecommerce products.

Most blog teams only need the first route. The others cover content the native export does not handle.

Publish blog articles with Export to Webflow

Export to Webflow is a direct integration built into Hypotenuse AI. You set it up once and export from the blog editor afterward. Hypotenuse AI documents it for blog articles only, so product descriptions need one of the other routes. You need a Webflow login with access to the site and a blog Collection to export into.

Connect your Webflow account

The connection lives on the Hypotenuse AI Integrations page and takes a single Webflow sign-in.

To connect the accounts:

1. Open Integrations in Hypotenuse AI and click Connect on the Webflow card.
2. Sign in to Webflow when prompted.
3. Return to Hypotenuse AI and select the workspace to link.

A CONNECTED label appears on the card once the link is active. Each user can connect one Webflow account. Within it, you can link multiple workspaces or several sites in one workspace.

Export an article to a Collection

Once connected, export runs from the blog editor and maps Hypotenuse AI content onto your Collection fields.

To export a post:

1. Open the finished article and click the arrow-up button at the top right of the editor.
2. Choose Export to Webflow.
3. If you linked more than one workspace or site, select the one to export to.
4. Select the blog post Collection.
5. Map each Webflow field on the left to Hypotenuse AI content on the right, then click Export.

The mapping screen offers these Hypotenuse AI values for your fields:

• Article title: This is the name of your article.
• Article content: This is the body of the blog post.
• Article outline: This holds the article's H2 headings.
• Article keywords: These are the keywords you entered at the title stage.
• Custom values: These let you type your own text for a field.

Hypotenuse AI's Webflow integration page says exports "Retain the original formatting and headings". Its pages disagree on whether an export goes live, though. That page says an exported post "appears on your site instantly", while the FAQ further down it says "exporting to Webflow does not publish them to your live site". The Blog Pro guide says you can "save posts as a draft or publish directly to your site".

Check the item's status in the CMS panel after your first export and plan your review step around what you see. A Collection List on your blog page renders the item once it is published.

Import bulk exports into the Webflow CMS

Bulk Actions in Hypotenuse AI export product descriptions as a CSV file. The bulk rewriter exports CSV or XLSX, and bulk blogs export as CSV. Meta titles and descriptions run through their own bulk workflow. The Webflow CMS imports CSV files into any Collection.

This route suits catalog teams who generate hundreds of descriptions at once, and it needs no API project. It also covers content the native export skips, such as product copy and metadata. The CMS import article covers the Webflow side in full.

Import a CSV into a Collection

Match your CSV headers to Collection field names before upload, and the importer maps them automatically.

You can upload a CSV file up to 4 MB, and your Site plan sets how many items you can import. If a row leaves a required reference field empty or unmatched, that item imports as a Draft. If the reference field is not required, the item is set to Staged for publish instead.

To import a file:

1. In Hypotenuse AI, import your products from CSV or XLSX, run Bulk Generate, then click Export to download the CSV.
2. Rename the columns to match your Collection field names. If you save the file from Microsoft Excel, choose the Comma Separated Values CSV type.
3. Open the CMS panel, choose the Collection, and click Import.
4. Drag in the CSV, confirm the header row, and click Continue.
5. Choose which locales to import into if prompted, click Continue, map the fields, then click Import.

Some field types need specific values in the CSV:

• Rich text: Any value is accepted, including HTML, though Webflow may remove elements, classes and attributes it does not support.
• Image: Each cell needs a direct image URL, and images imported this way get no responsive srcset or sizes variants.
• Multi-image: Separate the image URLs in a cell with semicolons.
• Option: Every unique value in the column becomes an option, up to 100 options per field.
• File: No CSV data maps to a File field, so upload those files by hand after the import.

If your CSV has only one column, add a comma directly after the header label so it imports correctly.

Paste Hypotenuse AI exports into rich text or Code Embed

For a one-off article, export it from Hypotenuse AI as DOCX and open it in Google Docs before you paste it into a Rich text element. Content pasted from a word processor like Google Docs keeps its formatting and links, and images import on paste. Press releases can also export as HTML, which you can drop into a Code Embed element on a static page.

To paste a document:

1. Click the arrow-up button in the Hypotenuse AI editor and choose DOCX, or HTML for a press release.
2. Open the DOCX in Google Docs rather than Word, copy the body and paste it into the rich text field.
3. For HTML, open the Add panel, add a Code Embed element, paste the markup, and save.

Adding a Code Embed requires a Core, Growth, Agency or Freelancer Workspace, or an active Site plan. Each embed holds up to 50,000 characters.

Build with Hypotenuse AI exports and the Webflow Data API

Hypotenuse AI describes its "Product Description Generator API" as a REST API that lets you "send product information and retrieve back product descriptions for your PDP". Its API page offers an API key on request and links no public endpoint reference, so the workflows below start from files you export from Hypotenuse AI instead of from Hypotenuse AI API calls.

On the Webflow side, the workflows call the Data API at https://api.webflow.com/v2 with an Authorization: Bearer token. They use the CMS Collection endpoints and the Ecommerce Products endpoints, and they suit ecommerce teams with a developer who runs the sync on their own server.

Create CMS items from exported articles

Use the Create Items endpoint to stage many posts in one request, including in Collections the native export does not reach. Export each article from Hypotenuse AI as TXT (markdown), convert it to an HTML string, and send it as the value of a RichText field. Items from this endpoint are created as drafts by default, so editors can review them before anything goes live.

To implement this:

1. Call GET /v2/collections/{collection_id} with the cms:read scope to read each field's slug, type and isRequired value.
2. Call POST /v2/collections/{collection_id}/items/insert with the cms:write scope, sending an items array whose fieldData holds name, slug and your RichText field. One request can create up to 100 items.
3. After review, call POST /v2/collections/{collection_id}/items/publish with the new item IDs in itemIds.

The request body for step 2 looks like this:

{
  "items": [
    {
      "isDraft": true,
      "fieldData": {
        "name": "How to size a snowboard",
        "slug": "how-to-size-a-snowboard",
        "<rich_text_field_slug>": "<h2>Start with your height</h2><p>...</p>"
      }
    }
  ]
}

To skip review, create items with POST /v2/collections/{collection_id}/items/live instead, which publishes them immediately. The API does not support code blocks in RichText fields and turns them into an empty string, so keep code samples out of articles you send this way. To alert reviewers when an item lands, register a collection_item_created webhook and verify each delivery against the x-webflow-signature header. That header is a SHA-256 HMAC of the x-webflow-timestamp value and the request body, joined by a colon.

Load bulk product descriptions into Webflow Ecommerce

Hypotenuse AI documents a direct export of product descriptions to Shopify, but no Webflow product sync. The bridge is its CSV export plus the Webflow Products endpoint.

To implement this:

1. In Hypotenuse AI, import products from CSV or XLSX (the sample template asks for product title, image source, brand and product features), run Bulk Generate, and export the CSV.
2. For each row, call POST /v2/sites/{site_id}/products with the ecommerce:write scope. Map the product title to fieldData.name and the description to fieldData.description, add a generated slug, and include the sku object for the default SKU (stock-keeping unit).
3. Set publishStatus to staging, the default, to hold the product for review, or to live to publish it. Use PATCH /v2/sites/{site_id}/products/{product_id} for later edits.

Leaving products in staging lets a merchandiser check the copy before it reaches the live store.

Update meta titles and descriptions in bulk

This workflow writes new meta titles and descriptions onto existing CMS Collection items.

To implement this:

1. Run the bulk meta workflow in Hypotenuse AI and export the file, which lists the webpage URL beside each generated meta title and description.
2. Call GET /v2/collections/{collection_id}/items and match each row to an item by the slug in its webpage URL.
3. Call PATCH /v2/collections/{collection_id}/items to update up to 100 items per request.

Since July 7, 2025, setting isDraft to true on a live item no longer unpublishes it. Use the dedicated unpublish endpoints when you need to take an item down, as the API changelog explains.

What you can build with the Hypotenuse AI Webflow integration

These builds pair Hypotenuse AI content with the Webflow CMS and Webflow Ecommerce.

• SEO blog pipeline: A B2B company drafts articles in the blog writer with SERP research and automatic linking. It exports each one into its Webflow posts Collection with title, content, outline and keywords mapped.
• Multi-client agency publishing: An agency links each client's workspace through one connected Webflow account and applies that client's brand voice before exporting posts. Hypotenuse AI's agencies page pitches writing in "10 or 100 clients’ brand voices and styles".
• Seasonal catalog launch: A fashion store loads a new season's descriptions from a Bulk Actions CSV into Webflow Ecommerce through the Products endpoint. Each product stays in staging for review before it goes live.
• Metadata refresh: An SEO specialist regenerates meta titles and descriptions in bulk and patches the results into PlainText fields on existing blog and category items, up to 100 items per request.

The last three builds run on your own code, so start with our CMS API walkthrough before you write it.

If you would rather have an AI agent manage those Collection items for you, connect the Webflow MCP server to your AI client. It can manage CMS items with bulk updates and publish or unpublish them.