A side-by-side look at the Grok MCP and Grok MCP MCP servers: transport, auth, tools, quality signals, and the config to install either one in Claude, Cursor, Windsurf or VS Code.
Verdict: choose Grok MCP for conversational ai work as a local process; choose Grok MCP for conversational ai. You can also run both side by side in one mcpServers block.
XAI_API_KEY, GROK_MCP_TIMEOUT, GROK_MCP_PATH_SECRET.grok_chat, grok_review, grok_consult.XAI_API_KEY.chat, chat_with_vision, chat_with_files.Rows where the two servers differ are marked.
| Spec | ||
|---|---|---|
| Category | Conversational AI | Conversational AI |
| Quality signal | 60/100 · Good | 52/100 · Good |
| Install command | Confirmed | Not confirmed yet |
| Transport | Local subprocess (stdio) | — |
| Auth | Not declared | API key required |
| Pricing | Not declared | BYOK (pay the provider) |
| Required env vars | XAI_API_KEYGROK_MCP_TIMEOUTGROK_MCP_PATH_SECRET | XAI_API_KEY |
| Verified owner | Community listing | Community listing |
| GitHub stars | 11 | 55 |
| Directory activity | 1 views 0 installs 0 upvotes | 3 views 0 installs 0 upvotes |
| Tools | 4 | 22 |
| Full listing | View Grok MCP | View Grok MCP |
grok_chatSend a one-shot prompt to xAI Grok and return its reply. Stateless — for multi-turn use grok_consult.grok_reviewHave Grok review a git diff. If no diff is provided, runs `git diff <base_ref>...HEAD` (default base: main). Returns a per-dimension review. Use format='json' for CI gating (verdict + scores + blockers).grok_consultContinue a conversation with Grok by replaying the full message history each call. Stateless on the server side — the caller owns the thread.grok_challengeAsk Grok to adversarially break a piece of code: edge cases, race conditions, security holes, adversarial inputs. Returns severity-ranked issues with reproductions.chatstandard chat completion with optional persistent history, multi-agent support,chat_with_visionanalyze local or remote images with a Grok vision model.chat_with_fileschat grounded on previously uploaded documents.stateful_chatcontinue a server-side stored conversation via `response_id`.retrieve_stateful_responsefetch a stored response by ID.delete_stateful_responsedelete a stored response by ID.web_searchautonomous web research with domain filters and citations.x_searchautonomous search over X (Twitter) posts, with handle and date filters.code_executorsolve tasks by running Python in a sandbox.grok_agentunified agent that mixes files, images, web search, X search, code execution,generate_imagecreate or edit images with Grok Imagine (multi-reference editing supported).generate_videotext-to-video, image-to-video, or video editing with Grok Imagine.Paste either block (or both) into your client config, e.g. claude_desktop_config.json or ~/.cursor/mcp.json. We only show a config when the install command is confirmed.
{
"mcpServers": {
"grok-mcp": {
"command": "npx",
"args": [
"-y",
"grok-cli-mcp"
],
"env": {
"XAI_API_KEY": "YOUR_XAI_API_KEY_HERE",
"GROK_MCP_TIMEOUT": "YOUR_GROK_MCP_TIMEOUT_HERE",
"GROK_MCP_PATH_SECRET": "YOUR_GROK_MCP_PATH_SECRET_HERE"
}
}
}
}
No confirmed install command for Grok MCP yet — follow the setup steps linked from its listing page rather than a guessed package name.