Gemini CLI MCP Server

Gemini CLI MCP server enabling AI agents to interact with Google's Gemini CLI tool — running Gemini AI queries, code generation, and analysis tasks through the Gemini CLI interface, integrating Google's Gemini models into agent-driven workflows where direct Gemini API access isn't available or where Gemini CLI-specific features are needed.

Evaluated Mar 06, 2026 (0d ago) vcurrent
Homepage ↗ Repo ↗ AI & Machine Learning gemini google-ai mcp-server llm ai-ml code-generation
⚙ Agent Friendliness
67
/ 100
Can an agent use this?
🔒 Security
78
/ 100
Is it safe for agents?
⚡ Reliability
63
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
62
Documentation
65
Error Messages
62
Auth Simplicity
80
Rate Limits
72

🔒 Security

TLS Enforcement
95
Auth Strength
78
Scope Granularity
68
Dep. Hygiene
65
Secret Handling
78

HTTPS to Google AI. API key via env var. Data sent to Google. Community MCP.

⚡ Reliability

Uptime/SLA
70
Version Stability
62
Breaking Changes
60
Error Recovery
60
AF Security Reliability

Best When

An agent workflow needs to invoke Google's Gemini CLI for tasks — particularly useful in multi-model setups where Claude handles orchestration and Gemini handles specific analysis.

Avoid When

You can use the Gemini API directly — CLI wrapper adds unnecessary latency and complexity.

Use Cases

  • • Running Gemini AI queries from multi-model comparison agents
  • • Using Gemini for code generation alongside other AI models from coding agents
  • • Integrating Gemini CLI's file analysis capabilities from research agents
  • • Building workflows that leverage both Claude and Gemini from orchestration agents
  • • Accessing Gemini's large context window for long document analysis
  • • Testing cross-model AI capabilities from evaluation agents

Not For

  • • Teams wanting direct Google Gemini API access (use google-generativeai SDK)
  • • Production high-volume Gemini usage (CLI has overhead vs direct API)
  • • Teams without Gemini API key or Google AI Studio access

Interface

REST API
No
GraphQL
No
gRPC
No
MCP Server
Yes ↗
SDK
No
Webhooks
No
OpenAPI Spec ↗

Authentication

Methods: api_key
OAuth: No Scopes: No

Google AI Studio API key or Google Cloud credentials required for Gemini CLI. API key set via GEMINI_API_KEY environment variable.

Pricing

Model: usage_based
Free tier: Yes
Requires CC: No

Gemini API has a free tier sufficient for development. Production use requires billing. Community MCP server is free.

Agent Metadata

Pagination
none
Idempotent
Full
Retry Guidance
Not documented

Known Gotchas

  • ⚠ Gemini CLI must be installed locally — not just the API key
  • ⚠ CLI process spawning adds latency vs direct Gemini API calls
  • ⚠ Gemini CLI version compatibility — CLI flags and behavior change with updates
  • ⚠ Rate limits apply per API key — shared keys deplete quickly in multi-agent scenarios
  • ⚠ Community MCP — Gemini CLI updates can break wrapper behavior
  • ⚠ Output parsing may fail for complex Gemini CLI output formats

Alternatives

mcp-server-openai mcp-anthropic-server mcp-vertex-ai

Full Evaluation Report

Detailed scoring breakdown, competitive positioning, security analysis, and improvement recommendations for Gemini CLI MCP Server.

$99
API endpoint ↗ Agent guide ↗ Report inaccuracy

Scores are editorial opinions as of 2026-03-06.

5178
Packages Evaluated
26151
Need Evaluation
173
Need Re-evaluation
♥
Community Powered