Gemini Media Mcp

general MCP Server

Unified Go MCP server for AI media generation via Google Gemini API and Vertex AI

Verified
generalgeneralgcp
8 views5 stars4 forksApache-2.0

Why This Matters

Discovered via list:awesome-mcp-servers-punkpeye and last synced 3mo ago.

Verified
Source
list:awesome-mcp-servers-punkpeye
Stars
5
Last synced
3mo ago
Install
Check source

Install

Install instructions not detected yet

Check the source repository for the latest setup steps.

View source instructions
27
Tools
0
Resources
0
Prompts
Standard I/O
Transport

Available Tools (27)

Tool

Description

Variable

Required

animate_image

Animate image into video (first frame)

MEDIA_OUTPUT_DIR

No

compose_images

Multi-reference image composition (up to 3)

get_config

Show current backend and configuration

lyria-3-clip-preview

30-second clips

generate_video

Generate video from text prompt (returns operation ID)

Tier

Model

generate_audio

Generate spoken audio from text (TTS)

edit_image

Edit existing image with text prompt

list_models

Show available models with capabilities and pricing

generate_image

Generate image from text prompt

generate_music

Generate AI music from text description (Lyria)

tts

`gemini-2.5-flash-preview-tts`

GOOGLE_API_KEY

Yes*

pro

`gemini-3-pro-image-preview`

GOOGLE_CLOUD_LOCATION

No

video_status

Check video generation progress

fast

`veo-3.1-fast-generate-preview`

download_video

Download completed video

GOOGLE_CLOUD_PROJECT

Yes*

extend_video

Chain video clips for longer content

Directory

Description

standard

`veo-3.1-generate-preview`

full

`lyria-3-pro-preview`

Token-based

The `generate_music` tool creates AI-generated music from text descriptions. Capabilities include: - **Genre and style** -- specify any genre, instruments, BPM, key/scale, mood - **Structure control** -- use tags like `[Verse]`, `[Chorus]`, `[Bridge]`, `[Intro]`, `[Outro]` - **Custom lyrics** -- include lyrics with section markers for vocal tracks - **Timestamp control** -- `[0:00 - 0:10] Intro: gentle piano...` for precise section timing - **Multi-language** -- prompt language determines output language - **High fidelity** -- 48kHz stereo MP3 output All generated music is watermarked with SynthID. **Example prompts:** ``` # Instrumental "A gentle acoustic guitar melody in C major, 90 BPM, calm and peaceful indie folk" # With structure "[Intro] Ambient synth pad, ethereal [Verse] Lo-fi hip-hop beat, mellow piano chords, vinyl crackle [Chorus] Uplifting, add strings and gentle drums [Outro] Fade out with reverb" # With lyrics "Upbeat pop song, 120 BPM, major key [Chorus] We're dancing in the light / Everything feels right / Under stars so bright tonight" ``` You can pass the tier name (`lite`, `fast`, `standard`, `nb2`, `pro`, `tts`, `clip`, `full`) or a raw model ID directly. ## MCP Client Configuration ### Claude Code Add to your Claude Code MCP settings (`~/.claude/settings.json` or project `.mcp.json`): ```json { "mcpServers": { "gemini-media": { "command": "gemini-media-mcp", "env": { "GOOGLE_API_KEY": "your-api-key", "MEDIA_OUTPUT_DIR": "/path/to/output" } } } } ``` Use either `GOOGLE_API_KEY` or `GEMINI_API_KEY` in the `env` block above; both are accepted. Or if building from source: ```json { "mcpServers": { "gemini-media": { "command": "/path/to/gemini-media-mcp", "env": { "GOOGLE_API_KEY": "your-api-key" } } } } ``` ## Companion Skills for Claude Code The `skills/` directory contains Claude Code skills that provide interactive workflows on top of the MCP tools. Each skill guides Claude through prompt engineering, model selection, and iterative refinement for a specific media type.