> ## Documentation Index
> Fetch the complete documentation index at: https://docs.keenable.ai/llms.txt
> Use this file to discover all available pages before exploring further.

> ## Agent Instructions
> Every page is listed in https://docs.keenable.ai/llms.txt. To pick a product for a job, read https://docs.keenable.ai/choose.md
> Setup steps for agents, covering the SDKs, the MCP server and the CLI: https://docs.keenable.ai/skill.md
> Keenable search and fetch work without an API key. To add them to code, use the official SDK for the project's language: https://docs.keenable.ai/sdks.md
> Without a key, call POST https://api.keenable.ai/v1/search/public or GET https://api.keenable.ai/v1/fetch/public with an X-Keenable-Title header that names your app.
> With a key, send it in the X-API-Key header and read it from the KEENABLE_API_KEY environment variable. Never ask a user to paste a key into chat. Every access method: https://docs.keenable.ai/auth.md
> To get an API key without the console, use device login: https://docs.keenable.ai/authentication.md#device-login
> MCP server: https://api.keenable.ai/mcp (no key, or X-API-Key; keyless Codex uses https://api.keenable.ai/mcp-public), or https://api.keenable.ai/mcp-oauth for OAuth sign-in. These docs are an MCP server at https://docs.keenable.ai/mcp and the OpenAPI spec is at https://docs.keenable.ai/api-reference/openapi.json
> Keenable searches and fetches public web pages. If a question needs no web information, skip the call.
> On 402 with an API key (code insufficient_credits), stop and ask a human to add credits; don't retry. On 429, wait the number of seconds in Retry-After, then retry. Before you ship, follow https://docs.keenable.ai/production.md
> Every error, with its cause and fix, is in https://docs.keenable.ai/errors.md and moving from another search API is covered in https://docs.keenable.ai/migrate.md
> Before you compare Keenable with another search API, read https://docs.keenable.ai/guides/evaluate-search-apis.md

# CowAgent web search provider

> Keenable is a built-in web search provider in CowAgent that can run without an API key.

*Search provider*

[CowAgent](https://github.com/zhayujie/CowAgent) is an open-source AI assistant that plans tasks, runs Skills, and works from a web console and the major IM platforms. Keenable is one of the backends of its built-in `web_search` tool. If you turn it on with an empty key field, searches go to Keenable's public endpoint, with no account needed.

→ [The Keenable provider in CowAgent](https://github.com/zhayujie/CowAgent/pull/3122)

> [!NOTE]
>
> Keenable ships with CowAgent **2.1.8** and later. Install CowAgent with the [quick start](https://docs.cowagent.ai/guide/quick-start), or bring an existing install up to date with `cow update` ([upgrade guide](https://docs.cowagent.ai/guide/upgrade)).

## Turn it on

**1. Open the search settings**

In the [web console](https://docs.cowagent.ai/channels/web) (`http://localhost:9899` by default), open **Config**, switch to the **Models** tab, and find the **Web Search** card.

**2. Add Keenable**

Click **Add provider** and pick **Keenable**. Leave **API Key** empty and click **Save**. The card now lists **Keenable · anonymous**.

**3. Ask the agent**

Ask something that needs the web. The agent calls its `web_search` tool, which answers from Keenable when no other search provider is set up.

The console saves this to `config.json` (`~/CowAgent/config.json` for a script install, `~/.cow/config.json` for the desktop app). To set it by hand, add this block and run `cow restart`:

```json theme={"theme":{"light":"github-light-high-contrast","dark":"dark-plus"}}
{
  "tools": {
    "web_search": {
      "keenable_anonymous": true
    }
  }
}
```

## Which provider answers

The search strategy defaults to `auto`, which uses the first configured provider in this order: DeepSeek, Bocha, Baidu Qianfan, Zhipu, LinkAI, AnySearch, Serply, Tavily, Outcrawl, Firecrawl, SpecStream, SearXNG, Keenable. Keenable comes last, so a provider you already set up keeps answering. With two or more providers configured, the agent can still pick Keenable for a single call. To always search with Keenable, set **Strategy** to **Pinned** and **Provider** to **Keenable** in the same card and click **Save**, or in `config.json`:

```json theme={"theme":{"light":"github-light-high-contrast","dark":"dark-plus"}}
{
  "tools": {
    "web_search": {
      "strategy": "fixed",
      "provider": "keenable",
      "keenable_anonymous": true
    }
  }
}
```

## What the tool returns

Each call returns up to 50 results (10 by default), each with a `title`, `url`, `snippet`, and `datePublished`. Snippets run about 300 characters, or about 1,000 when the agent asks for summaries. The tool's `freshness` filter (`oneDay`, `oneWeek`, `oneMonth`, `oneYear`, or a range such as `2025-01-01..2025-02-01`) limits results by publish date.

## Use your API key

Paste your key (`keen_<your_key>`, from the [console](https://app.keenable.ai/console)) into **API Key** in the same dialog, or set `tools.web_search.keenable_api_key` in `config.json`. If that value is empty, CowAgent reads the `KEENABLE_API_KEY` environment variable instead. With a key, Keenable counts as configured without `keenable_anonymous`.

> [!NOTE]
>
> Without a key, searches run on the shared public tier at [lower rate limits](https://docs.keenable.ai/rate-limits.md). A key moves them to the authenticated endpoint and your organization's limits, and removes the hourly cap.

## Related pages

> [card] [LibreChat](https://docs.keenable.ai/integrations/librechat.md)
>
> Keenable is the keyless search provider and content scraper built into LibreChat's web search.

> [card] [Web search in Gemini and Gemini CLI](https://docs.keenable.ai/integrations/gemini.md)
>
> Add Keenable to the Gemini app and Gemini CLI, and keep Gemini CLI's own web search from competing with it.

> [card] [tgpt](https://docs.keenable.ai/integrations/tgpt.md)
>
> Use Keenable as the web search provider in tgpt, the terminal AI chat, with SEARCH\_PROVIDER=keenable. No API key needed, in tgpt 2.16.0 and later.

> [card] [Connect your agent (MCP)](https://docs.keenable.ai/mcp-server.md)
>
> Connect the Keenable MCP server to Claude Code, Cursor, Codex and other clients for web search and fetch with no key. Sign in for higher limits and Monitors.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.