# Desearch MCP server

> Search the web and X with source links, and extract clean content from any URL for AI agents.

- Listing: https://mcp.tc/i/desearch
- Connect: use the server's own URL `https://mcp.desearch.ai/mcp` (OAuth sign-in at the server); clients connect to it directly. The listing link is a page, not an MCP endpoint.
- Type: remote (Streamable HTTP)
- Auth: OAuth sign-in
- Category: [Search & Web Data](https://mcp.tc/c/search-web)
- Vendor: Desearch
- Homepage: <https://www.desearch.ai/docs/guide/sdk/mcp>
- Docs: <https://www.desearch.ai/docs/guide/sdk/mcp>

## About

Desearch is a search API for AI agents that returns fresh, source-linked results from the web and X (Twitter). The same service also extracts clean content from URLs and covers public TikTok, Instagram and Facebook endpoints.

The MCP server runs as a hosted streamable HTTP endpoint at https://mcp.desearch.ai/mcp. Access needs Desearch credentials; the vendor's API key is issued from its console.

## What it can do

- Search the web with source links
- Search X (Twitter) posts
- Extract clean content from any URL
- Get answers grounded in returned sources

## Tools (39)

- `ai-search`: AI Search over X (Twitter) posts and web (from the Desearch index), with an optional AI summary. (read-only)
- `x-search`: Search X (Twitter) using Desearch AI. (read-only)
- `web-search`: SERP-style web search using Desearch. (read-only)
- `web-links-search`: Deprecated: use search-api. (read-only)
- `x-links-search`: AI search for X (Twitter) post links using Desearch. (read-only)
- `x-posts-by-urls`: Fetch full X (Twitter) posts for a list of post URLs. (read-only)
- `x-post-by-id`: Fetch a single X (Twitter) post by its ID. (read-only)
- `x-posts-by-user`: Search X (Twitter) posts by a specific user, with an optional keyword query. (read-only)
- `x-post-retweeters`: List users who retweeted an X (Twitter) post. (read-only)
- `x-user-posts`: Retrieve a user's X (Twitter) timeline posts by username. (read-only)
- `x-user-replies`: Fetch posts and replies by an X (Twitter) user, with an optional keyword query. (read-only)
- `x-post-replies`: Fetch replies to an X (Twitter) post, with an optional keyword query. (read-only)
- `extract`: Extract a public URL and return its content as plain text, markdown or HTML using Desearch. (read-only)
- `web-crawl`: Crawl a public URL and return its content as plain text or HTML on the legacy Desearch /web/crawl route. (read-only)
- `x-trends`: Retrieve trending topics on X (Twitter) for a location by its WOEID using Desearch. (read-only)
- `search-api`: Web search on the Desearch index (POST /search). (read-only)
- `tiktok`: Read public TikTok data with Desearch. (read-only)
- `facebook`: Read public Facebook data with Desearch. (read-only)
- `instagram`: Read public Instagram data with Desearch. (read-only)
- `youtube`: Read public YouTube data with Desearch. (read-only)
- `reddit`: Read public Reddit data with Desearch. (read-only)
- `threads`: Read public Threads data with Desearch. (read-only)
- `search`: Search the public web and return titles and links. (read-only)
- `fetch`: Read a public page as text. (read-only)
- `people-search`: Start a paid job billed to the user's Desearch balance: find people on X, YouTube, LinkedIn, or TikTok who match a plain-language description.
- `people-search-status`: Status, progress, and cost of one People Search task, plus a page of the people found so far: name, handle, url, platform, audience, location, fit score, why they matched, contacts, and links. (read-only)
- `people-search-pause`: Pause a People Search task and return its updated status.
- `people-search-resume`: Resume a paused People Search task and return its updated status.
- `people-search-cancel`: Cancel a People Search task. (can modify or delete data)
- `people-search-tasks`: Your People Search tasks, newest first, with status, counts, cost, why each ended search stopped (completion\_reason, coverage\_limited), and the metadata labels set on people-search. (read-only)
- `people-profiles`: Everyone your own People Searches found on one platform, one entry per person: their best-scoring result (name, handle, url, audience, location, fit score, why they matched, public contacts, and links) and the search… (read-only)
- `people-search-export`: The people your own People Searches found, as CSV or JSON Lines text: one search (task\_id), or everyone found on one platform (platform, with the same q, sort, and filters as people-profiles). (read-only)
- `company-search`: Start a paid job billed to the user's Desearch balance: find companies on the web that match a plain-language description, and collect each one's public contact data.
- `company-enrich`: Start a paid job billed to the user's Desearch balance: collect public contact data for companies you already know, by domain.
- `company-search-status`: Status, progress, and cost of one Company Search or Company Enrich task, plus a page of the companies delivered so far: domain, name, url, location, fit score (null for Company Enrich) and why it matched, primary\_email… (read-only)
- `company-search-pause`: Pause a Company Search or Company Enrich task and return its updated status.
- `company-search-resume`: Resume a paused Company Search or Company Enrich task and return its updated status.
- `company-search-cancel`: Cancel a Company Search or Company Enrich task. (can modify or delete data)
- `company-search-tasks`: Your Company Search and Company Enrich tasks, newest first, with status, counts, cost, why each ended task stopped (completion\_reason, coverage\_limited), max\_cost\_usd and continue\_from when a search set them, and the… (read-only)

## Example prompts

- "Search the web for recent news on Bittensor subnet 22 and cite sources."
- "Find recent X posts about the latest Python release."
- "Extract the main content from this article URL."
- "Research current approaches to vector database indexing with links."

## Install

### Claude Code

1. Run this in a terminal, in your project folder:

```bash
claude mcp add --transport http desearch https://mcp.desearch.ai/mcp
```

2. Start Claude Code, type `/mcp`, pick **desearch** and choose **Authenticate**. A browser window opens for the Desearch sign-in.

Add `--scope user` to make it available in every project, not just this one.

### Claude Desktop

1. Open **Customize → Connectors**, click **+ Add**, then **Add custom connector**.

2. Name it **Desearch** and paste this URL:

```url
https://mcp.desearch.ai/mcp
```

3. Click **Add**, then **Connect**, and sign in when Desearch asks.

Claude Desktop’s JSON config file only starts local servers. Remote servers go through Connectors, and connectors you add on claude.ai show up here too.

### claude.ai

1. Open the connector form on claude.ai. This button fills in the name and URL for you:

[Add to claude.ai](<https://claude.ai/customize/connectors?modal=add-custom-connector&connectorName=Desearch&connectorUrl=https%3A%2F%2Fmcp.desearch.ai%2Fmcp>) (opens connector settings)

2. Check that the URL reads `https://mcp.desearch.ai/mcp` and click **Add**.

3. Click **Connect** and sign in when Desearch asks.

Free plans allow one custom connector. On Team and Enterprise plans an owner adds it under **Organization settings → Connectors**.

### ChatGPT

1. On chatgpt.com, go to **Plugins**, select **+**, then **Add custom MCP server**.

2. Enter a name for the server.

3. Under **Connection**, paste `https://mcp.desearch.ai/mcp` as the server URL, then for authentication choose **OAuth**. ChatGPT sends you to Desearch to sign in.

4. Check **I understand and want to continue**, then select **Create as a plugin**.

Whether you can add a custom MCP server depends on your ChatGPT plan and your workspace permissions (in Business workspaces, admins and owners can).

### Cursor

[Add to Cursor](<https://cursor.com/install-mcp?name=desearch&config=eyJ1cmwiOiJodHRwczovL21jcC5kZXNlYXJjaC5haS9tY3AifQ%3D%3D>) (opens Cursor)

Or add it by hand to `~/.cursor/mcp.json` (all projects) or `.cursor/mcp.json` (this project):

`mcp.json`:

```json
{
  "mcpServers": {
    "desearch": {
      "url": "https://mcp.desearch.ai/mcp"
    }
  }
}
```

Cursor shows **Needs authentication** next to the server. Click the button beside it to sign in.

### VS Code

[Install in VS Code](<https://vscode.dev/redirect/mcp/install?name=desearch&config=%7B%22type%22%3A%22http%22%2C%22url%22%3A%22https%3A%2F%2Fmcp.desearch.ai%2Fmcp%22%7D>) (opens VS Code)

Or from a terminal:

```bash
code --add-mcp '{"name":"desearch","type":"http","url":"https://mcp.desearch.ai/mcp"}'
```

Or commit it to the repo in `.vscode/mcp.json`:

`.vscode/mcp.json`:

```json
{
  "servers": {
    "desearch": {
      "type": "http",
      "url": "https://mcp.desearch.ai/mcp"
    }
  }
}
```

VS Code asks you to sign in the first time the server starts.

### Devin Desktop

1. Add it to `~/.config/devin/mcp_config.json` (macOS and Linux) or `%APPDATA%\devin\mcp_config.json` (Windows):

`mcp_config.json`:

```json
{
  "mcpServers": {
    "desearch": {
      "serverUrl": "https://mcp.desearch.ai/mcp"
    }
  }
}
```

2. Refresh the MCP server list in Cascade and sign in when asked.

Devin Desktop is the new name for Windsurf. It reads `serverUrl` (or `url`) for remote servers.

### Codex

```bash
codex mcp add desearch --url https://mcp.desearch.ai/mcp
codex mcp login desearch
```

Or edit `~/.codex/config.toml` directly:

`config.toml`:

```toml
[mcp_servers.desearch]
url = "https://mcp.desearch.ai/mcp"
```

### Gemini CLI

```bash
gemini mcp add --transport http desearch https://mcp.desearch.ai/mcp
```

Then, inside Gemini CLI, run `/mcp auth desearch` to sign in.

This adds it to the current project. Add `-s user` to use it everywhere.

### Any client

Most clients accept this shape. Some name the URL field differently: `serverUrl` in Devin Desktop, `httpUrl` in Gemini CLI’s settings file.

```json
{
  "mcpServers": {
    "desearch": {
      "type": "http",
      "url": "https://mcp.desearch.ai/mcp"
    }
  }
}
```

Zed puts servers under `context_servers` in its settings. Cline needs `"type": "streamableHttp"`, or it assumes SSE.

Client only starts local servers? Bridge it with `npx -y mcp-remote https://mcp.desearch.ai/mcp`.

## Recent changes

- 2026-11-17 17:51 UTC: New version 0.11.0 (was 0.10.0), reported by the server. Source: automatic check.
- 2026-11-17 17:51 UTC: 3 tools changed. Server instructions changed. Source: automatic check. Worth a look: new server instructions. Changed: ai-search, search-api, web-links-search.
- 2026-11-13 14:51 UTC: New version 0.10.0 (was 0.9.0), reported by the server. Source: automatic check.
- 2026-11-05 14:53 UTC: New version 0.9.0 (was 0.8.0), reported by the server. Source: automatic check.
- 2026-11-05 14:53 UTC: 4 tools changed. Server instructions changed. Source: automatic check. Worth a look: new server instructions. Changed: company-search, company-search-cancel, company-search-status, company-search-tasks.

More on the History tab, with descriptions before and after for the newest entries: https://mcp.tc/i/desearch#history (Atom feed: https://mcp.tc/i/desearch/history.xml). It records what mcp.tc's anonymous checks saw, every 3 hours where the tool list can be read without sign-in; it is not a security review.

## Details

- Server version: 0.11.1
- MCP protocol version: 2025-11-25
- Last checked: 2026-11-18 (reachable, asks for credentials)
- Listed: 2026-10-27
- Updated: 2026-11-17

---
Source: https://mcp.tc/i/desearch (mcp.tc is an independent directory, not affiliated with this server's publisher). Corrections: https://mcp.tc/report
