> ## Documentation Index
> Fetch the complete documentation index at: https://docs.bkmark.it/llms.txt
> Use this file to discover all available pages before exploring further.
# AI Auto-Tagging, Content Summaries, and Smart Search
> Let AI handle the busywork: auto-tag bookmarks as you save them, get group suggestions, summarize pages in seconds, and search by meaning on Pro and Team.
Bkmark's AI features take the manual effort out of organizing your library. Instead of tagging every bookmark yourself or writing elaborate search queries, you can let AI suggest the right tags, recommend a group, summarize a page, and expand your search terms — all automatically.
All AI features require a **Pro** or **Team** plan. If you're on the Free plan, upgrade from **Settings → Billing** to enable them.
***
## Auto-tagging
When you save a new bookmark, Bkmark's AI analyzes the page title, description, and URL and suggests up to **five relevant tags** — each with a confidence score so you know how strongly the AI recommends it.
### How it works
You don't need to do anything to trigger auto-tagging; it runs automatically every time you create a bookmark. After the bookmark is saved, the suggested tags appear in the bookmark editor so you can review and accept them with one click.
Each suggestion includes a confidence score so you know how strongly the AI recommends it. For example:
```json theme={null}
{
"name": "machine-learning",
"confidence": 0.92
}
```
A confidence score close to `1.0` means the AI is very sure the tag fits. Scores below `0.5` indicate a weaker match — use your judgment on those.
### Accepting or dismissing suggestions
Create a link, note, code snippet, or image bookmark as usual.
After saving, an **AI Suggestions** panel appears showing up to five recommended tags with their confidence scores.
Click a suggested tag to add it, or click **Accept all** to apply every suggestion at once. Dismiss any tag you don't want by clicking the **✕** next to it.
You can always add or remove tags manually after accepting suggestions. AI recommendations are a starting point, not a final answer.
***
## Group suggestion
Alongside tag suggestions, Bkmark's AI looks at your existing groups and recommends the **single best-fitting group** for each new bookmark. This is particularly useful when you have a well-organized group hierarchy and want to maintain it without manually triaging every new save.
### How it works
When you save a bookmark, the AI compares the content against the names of all your existing groups and returns the one that fits best. If none of your groups are a strong match, the AI returns no suggestion rather than force a poor fit.
### Using the suggestion
The group suggestion appears in the same **AI Suggestions** panel as your tag suggestions. Click **Add to group** to accept it, or ignore it and assign the bookmark manually as usual.
Group suggestion only considers groups you've already created. If you haven't set up any groups yet, this feature has nothing to recommend — create a few groups first and the suggestions will start appearing automatically.
***
## Content summarization
Summarization lets you generate a **2–3 sentence summary** of any link bookmark directly inside Bkmark, without having to open the page. This is great for quickly deciding whether a saved article is worth reading or for refreshing your memory on something you saved weeks ago.
### How to summarize a bookmark
Click on any link bookmark to open its detail view.
Press the **Summarize** button in the bookmark toolbar.
Bkmark fetches the page content and returns a 2–3 sentence summary in seconds. The summary is displayed beneath the bookmark description.
### Rate limit
Summarization is rate-limited to **10 requests per day** per account. The counter resets at midnight UTC. If you hit the limit, you'll see a message telling you when it resets.
The rate limit applies across all bookmarks in your account, so use summarization on the pages where a quick overview would be most valuable — long articles, documentation pages, or anything you're not sure is worth your time.
***
## AI-enhanced search
On Pro and Team plans, Bkmark enhances your search queries before running them. When you type a search term, the AI **rewrites and expands your query** into 5–10 related keywords and synonyms. The expanded keyword set is then used to find bookmarks that are conceptually related to what you're looking for — not just bookmarks that happen to use your exact words.
### Example
You search for: `"fast api calls"`
The AI expands this to keywords like: `api`, `performance`, `latency`, `http`, `optimization`, `caching`, `async`, `network`, `requests`, `throughput`
Your results include bookmarks about HTTP performance, caching strategies, asynchronous patterns, and network optimization — even if those bookmarks never mention "fast api calls" directly.
### How to use it
AI-enhanced search is **on by default** for Pro and Team users. Just search as you normally would — the query expansion happens automatically in the background. You don't need to change how you search; Bkmark handles the expansion for you.
If you want to find an exact phrase or a specific URL, wrap your query in **quotes** (e.g., `"react server components"`) to bypass expansion and search for the literal string instead.
***
## Feature availability at a glance
* Full-text search
* Manual tagging
* Manual group assignment
* Everything in Free
* Auto-tagging with confidence scores
* Group suggestions
* Content summarization (10/day)
* AI-enhanced search
* Everything in Pro
* Shared team workspaces
* Unlimited Slack channel mappings