Skip to main content
Search provider in Qwen Audio Agent Keenable is a built-in web search provider in Qwen Audio Agent, the realtime voice assistant. When you ask it to search for something, the assistant’s web_search tool queries Keenable, and the titles, snippets and links enter the voice model’s context as sources. It works without an API key. → Web search in Qwen Audio Agent’s docs

Install

The Keenable preset is in Qwen Audio Agent 2.1.0 and later. Install or update it from npm. It needs Node.js 22.22.2 or later, or 24.15.0 or later:

Select Keenable

Open your configuration:
Add the provider next to your voice model key:
Restart the Gateway (qwenaudio), then open the TUI with qwenaudio tui or the browser UI with qwenaudio webui. Ask “Search for today’s weather in Hangzhou” and the answer comes back with source links underneath.

How the preset works

The preset connects to the Keenable MCP server at https://api.keenable.ai/mcp and calls its search_web_pages tool. It always uses that endpoint and tool, and ignores the QWEN_AUDIO_WEB_SEARCH_MCP_* settings, which are for custom MCP search services. KEENABLE_API_KEY is optional. With a key, calls use your organization’s limits and are billed to its monthly allowance, then its balance.
Without a key, searches run on the shared public tier at lower rate limits. Uncomment KEENABLE_API_KEY in your configuration to use your organization’s limits. See API keys and the MCP server reference.

Pipecat

KeenableWebSearch gives Pipecat voice agents low-latency web search and page fetch.

Tavus

Give a Tavus PAL live web search with one API tool that calls Keenable, so it can answer questions about current events mid-conversation.

Search quickstart

Give your agent ranked web pages with excerpts ready to cite. Try it with no API key. Code, cost per mode, filters by site, date and domain, and gotchas.