Lets AI assistants call the Google Gemini CLI to analyze large files and codebases using Gemini's large context window.
Setup
Choose your client, copy the config, then fill in any required environment variables.
claude mcp add gemini-tool -- npx -y gemini-mcp-tool~/.codex/config.toml
[mcp_servers.gemini-tool]
command = "npx"
args = ["-y", "gemini-mcp-tool"]
# codex mcp add gemini-tool -- npx -y gemini-mcp-tool~/.cursor/mcp.json
{
"mcpServers": {
"gemini-tool": {
"command": "npx",
"args": [
"-y",
"gemini-mcp-tool"
]
}
}
}.vscode/mcp.json
{
"servers": {
"gemini-tool": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"gemini-mcp-tool"
]
}
}
}~/.gemini/settings.json
{
"mcpServers": {
"gemini-tool": {
"command": "npx",
"args": [
"-y",
"gemini-mcp-tool"
]
}
}
}claude_desktop_config.json
{
"mcpServers": {
"gemini-tool": {
"command": "npx",
"args": [
"-y",
"gemini-mcp-tool"
]
}
}
}.trae/mcp.json
{
"mcpServers": {
"gemini-tool": {
"command": "npx",
"args": [
"-y",
"gemini-mcp-tool"
]
}
}
}