Skip to main content
Mercury is served through an OpenAI-compatible API, so most tools that let you configure a custom “OpenAI Compatible” provider work with Inception out of the box. In almost every integration you’ll supply the same three values:

Base URL

https://api.inceptionlabs.ai/v1

API Key

Your Inception API key

Model

mercury-3 (chat) or mercury-edit-2 (code)

​
Coding assistants & editors

Cursor

Add Mercury 3 as a custom model in Cursor.

Zed

Enable Mercury Edit 2 for Zed’s Edit Predictions.

Kilo Code

Configure the Kilo Code VS Code extension.

OpenCode

Use Mercury from the OpenCode terminal client.

OpenClaw

Connect the OpenClaw personal assistant.

​
Frameworks & voice platforms

LangChain

Call Mercury 3 through langchain-openai.

ElevenLabs

Wire Mercury 3 into the ElevenLabs Agents platform.

Vapi

Build a low-latency voice agent in Vapi.
Using a tool that isn’t listed here? Any OpenAI-compatible client works — point it at the base URL above and use your Inception API key. See the third-party library examples for AISuite, LiteLLM, the OpenAI SDK, and the Vercel AI SDK.