​

Mem API Documentation

​

Overview

The Mem API provides two primary endpoints for creating content in Mem:

• Create Note (/v2/notes): For precise control over structured note creation
• Mem It (/v2/mem-it): For intelligent processing of raw, unstructured content

​

Choosing the Right Endpoint

​

Common Use Cases

​

Create Note

• Task management integrations
• Calendar event to note conversion
• CRM integrations
• Structured data from forms
• Templates with consistent formatting

​

Mem It

• Web page clipping
• Email processing
• Meeting transcript analysis
• Chat/message archiving
• Quick capture without formatting

​

Create Note API

Create structured, formatted notes with precise control over organization and formatting.

Endpoint: POST https://api.mem.ai/v2/notes

​

Request Parameters

​

Required Parameters

​

content

Markdown formatted string representing the note content. The first line is automatically interpreted as the title.

Example:

# Meeting with Product Team

Discussion Topics:
- Roadmap updates
- Feature prioritization

Action Items:
- Schedule follow-up with design

​

Optional Parameters

​

add_to_collections

Array of strings representing collection titles or IDs to add the note to. If a collection doesn’t exist, a new one will be created.

Example:

"add_to_collections": ["Meetings", "Product", "Q2 Planning"]

Example with IDs:

"add_to_collections": ["8a7b6c5d-4e3f-2a1b-0c9d-8e7f6a5b4c3d", "Marketing"]

​

auto_organize

Boolean or string that controls automatic organization. If true, Mem will automatically organize the note into relevant collections. If a string is provided, it will be used as custom organization instructions.

Default: true

Examples:

"auto_organize": true

"auto_organize": "Add this note to exactly one of my Marketing, Sales, or Recruiting collections"

​

apply_template

String representing a template title or ID to apply to the note content.

Example:

"apply_template": "Meeting Notes"

Example with ID:

"apply_template": "8a7b6c5d-4e3f-2a1b-0c9d-8e7f6a5b4c3d"

​

auto_format

Boolean or string that controls automatic formatting. If true, Mem will automatically format the note content. If a string is provided, it will be used as custom formatting instructions.

Default: true

Examples:

"auto_format": true

"auto_format": "Use bullet points throughout and highlight key action items"

​

created_at

ISO 8601 datetime string to set as the note’s creation time.

Default: current time

Example:

"created_at": "2025-04-01T14:30:45Z"

​

updated_at

ISO 8601 datetime string to set as the note’s last update time.

Default: same as created_at

Example:

"updated_at": "2025-04-02T09:15:22Z"

​

Example Requests

​

Basic Note Creation

curl --request POST \
  --url https://api.mem.ai/v2/notes \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{
    "content": "# Quick Thought\n\nNeed to follow up with marketing team about Q2 campaign."
  }'

​

Using a Template with Collections

curl --request POST \
  --url https://api.mem.ai/v2/notes \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{
    "content": "# Meeting with Product Team\n\nAttendees: Alex, Jamie, Taylor\n\nDiscussion: Roadmap planning",
    "add_to_collections": ["Meetings", "Product"],
    "apply_template": "Meeting Notes",
    "created_at": "2025-04-01T14:30:45Z"
  }'

​

Custom Organization Instructions

curl --request POST \
  --url https://api.mem.ai/v2/notes \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{
    "content": "# Sales Call with Acme Corp\n\nContact: John Smith (john@acme.com)\n\nInterested in enterprise plan. Follow up next week.",
    "auto_organize": "Add this to both my Sales and Follow-ups collections, and create a new Acme Corp collection if it doesn't exist",
    "auto_format": "Format as a sales call summary with clear follow-up tasks"
  }'

​

Response

A successful request returns a 201 Created status and the newly created note object:

{
  "id": "d4e78650-b159-47f2-a634-2c92f980c88d",
  "content": "# Meeting with Product Team\n\nAttendees: Alex, Jamie, Taylor\n\nDiscussion: Roadmap planning\n\nAction Items:\n- Schedule follow-up with design\n- Share Q2 priorities document\n- Update project timeline",
  "collections": [
    {
      "id": "8a7b6c5d-4e3f-2a1b-0c9d-8e7f6a5b4c3d",
      "title": "Meetings"
    },
    {
      "id": "1a2b3c4d-5e6f-7a8b-9c0d-1e2f3a4b5c6d",
      "title": "Product"
    }
  ],
  "created_at": "2025-04-01T14:30:45Z",
  "updated_at": "2025-04-01T14:30:45Z",
  "url": "https://mem.ai/m/d4e78650-b159-47f2-a634-2c92f980c88d"
}

​

Mem It API

The intelligent way to remember anything - just throw content at Mem and let it handle the rest.

Endpoint: POST https://api.mem.ai/v2/mem-it

​

Request Parameters

​

Core Parameters

​

input (required)

Any raw content you want to remember - web pages, emails, transcripts, articles, or simple text.

Examples:

• HTML from a webpage
• Email text (including headers)
• Meeting transcript
• Text from a document
• Simple notes or thoughts

​

instructions

Optional guidance on how you want this information processed or remembered.

Examples:

• “Extract the key ideas and action items from this meeting”
• “Save this recipe with just the ingredients and steps”
• “Add this to my reading list and extract author and genre”

​

context

Additional background information that helps Mem understand how this input relates to your existing knowledge.

Examples:

• “This is related to my Project Alpha research”
• “This is from the quarterly planning meeting”
• “This is a follow-up to our discussion yesterday”

​

timestamp

When this information was encountered or created (ISO 8601 datetime).

Default: Current time

Example:

"timestamp": "2025-04-01T14:30:45Z"

​

Source Attribution Parameters

Mem It lets you preserve a reference to the original content, which will appear as an attachment chip on any notes created from this input.

​

source_title

A display name for the original content. This will appear in the attachment chip’s label.

Examples:

• “Meeting Transcript - Product Team (April 2)”
• “Bloomberg Article: Tech Trends 2025”
• “Email from Jamie - Project Update”

​

source_url

Optional link to the original content. When provided, the attachment chip becomes clickable.

Examples:

• “https://news.example.com/tech-trends-2025”
• “outlook://message/AAMkAGNhOWY3…” (for email protocols)
• “slack://channel/C01234ABCDE/p1234567890123456” (for app-specific deep links)

​

source_kind

Optional string that describes the kind of source. Mem will display an appropriate icon based on the provided value.

Common values:

• “webpage”
• “email”
• “transcript”
• “document”
• “message”
• “note”
• “calendar”
• “task”

You can provide any string value, and Mem will map it to the closest matching icon. If no match is found, a generic source icon will be used.

​

Example Requests

​

Remembering a Web Article

curl --request POST \
  --url https://api.mem.ai/v2/mem-it \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{
    "input": "<!DOCTYPE html><html><head><title>AI Trends 2025</title>...</html>",
    "instructions": "Extract the key findings and save as a research note",
    "source_title": "AI Trends 2025 - TechInsights",
    "source_url": "https://techinsights.com/ai-trends-2025",
    "source_kind": "webpage"
  }'

​

Remembering a Meeting Transcript

curl --request POST \
  --url https://api.mem.ai/v2/mem-it \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{
    "input": "[08:47:19] Kevin: Hello...[09:15:22] Meeting ended.",
    "instructions": "Create a summary with action items and decisions",
    "context": "Weekly product planning meeting",
    "timestamp": "2025-04-01T08:47:19Z",
    "source_title": "Product Planning - April 1",
    "source_kind": "transcript"
  }'

​

Quick Recommendation Capture

curl --request POST \
  --url https://api.mem.ai/v2/mem-it \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '{
    "input": "Check out Atomic Habits by James Clear - recommended by Sarah",
    "instructions": "Add to my reading list"
  }'

​

Response

A successful request returns a 201 Created status and information about the created content:

{
  "operation_id": "mem-it-8a7b6c5d-4e3f-2a1b-0c9d-8e7f6a5b4c3d",
  "created_notes": [
    {
      "id": "d4e78650-b159-47f2-a634-2c92f980c88d",
      "title": "AI Trends 2025: Key Findings",
      "url": "https://mem.ai/m/d4e78650-b159-47f2-a634-2c92f980c88d"
    }
  ],
  "collections": [
    {
      "id": "1a2b3c4d-5e6f-7a8b-9c0d-1e2f3a4b5c6d",
      "title": "Research Notes"
    }
  ],
  "status": "complete",
  "processing_summary": "Extracted key findings from article and created a research note"
}

​

Understanding Source Attribution

When you provide source information, Mem creates an attachment chip that preserves the connection to the original content. This creates a powerful experience where:

1. Users can see where information came from
2. The original unprocessed content remains accessible
3. Multiple notes created from the same source maintain a visual relationship

This is especially useful for applications like web clippers, email processors, or transcript analyzers, where the final notes may contain only a portion or transformation of the original content.

​

Authentication

All API requests require authentication using a Bearer token in the Authorization header:

Authorization: Bearer <your_api_token>

To obtain an API token, visit the API section in your Mem settings.

​

Rate Limits

• Free tier: 25 requests per day while in Alpha

When rate limits are exceeded, the API will return a 429 status code.

​

Error Handling

The API uses standard HTTP status codes to indicate success or failure:

• 2xx: Success
• 400: Bad request (invalid parameters)
• 401: Unauthorized (invalid token)
• 404: Resource not found
• 429: Rate limit exceeded
• 500: Server error

Error responses include a JSON body with details:

{
  "error": {
    "code": "invalid_parameter",
    "message": "The parameter 'content' is required",
    "param": "content"
  }
}

​

Best Practices

1. Choose the right endpoint based on your use case and content format
2. Use descriptive titles in your content or source_title
3. Leverage the auto-organization and auto-formatting features for best results
4. Provide context when using Mem It for better content understanding
5. Include source information when capturing external content
6. Set accurate timestamps for historical content