Name: Gemini Image
Author: JimothySnicket

Gemini Image MCP

Google Gemini image generation, editing, and local processing via MCP

Grade A Local · stdio API key 2 tools v0.6.2 · 16 Sep 2026 MIT

No reviews yet by JimothySnicketnpm 8,745 Runs on your machine, nothing hosted

96/100

trust score

downloads a day

local

runs on your machine

downloads in 30d

What it does

Google Gemini image generation, editing, and local processing via MCP

Quickstart

# 1 — add it to Claude Code
claude mcp add --scope user --env 'GEMINI_API_KEY=${GEMINI_API_KEY}' --transport stdio jimothysnicket-gemini-image-mcp '--' npx -y @jimothy-snicket/gemini-image-mcp

# 2 — ask your agent for something these tools do
#   generate_image, process_image

It needs GEMINI_API_KEY: this reads it from your environment, so export it first.

Gemini Image costs nothing on mcprush: there is no plan to choose here, no cap to set and nothing mcprush can bill you for.

Publishernot claimed

JimothySnicket1 server

1 category

All 1 server

Collected from a public index. Nobody has claimed this account, so nothing here was written by its author — claim it if it is yours.

Installation

Where are you running it?

Every route below installs the same thing and ends at the same approval screen. This one runs on your machine: your client starts Gemini Image as a process under your own user, with your files and your network, so the tool surface below is what it can do to your computer rather than to a server somewhere else. Its scan result is on the Security tab — read the surface before you approve it.

Claude Codedeclared by this listing

This is a public server: you run it yourself and this marketplace is not in the path. You need Node.js — the current LTS (24) is the safe choice; npx comes with it. Claude Code registers it in one command; --scope user makes it available in every project. Claude Code reads GEMINI_API_KEY (Google Gemini API key from https://aistudio.google.com/apikey) from your environment each time it starts the server, so the value stays out of its config: export it in your shell profile (macOS, Linux), or run setx GEMINI_API_KEY … and open a new terminal (Windows). The '--' is in quotes so that PowerShell passes it on.

claude mcp add --scope user --env 'GEMINI_API_KEY=${GEMINI_API_KEY}' --transport stdio jimothysnicket-gemini-image-mcp '--' npx -y @jimothy-snicket/gemini-image-mcp

Reconnect, or start a new session, and the tools appear in the model’s tool list.

Free on mcprush, and it runs from its publisher’s own package: this marketplace is not in the path, so there is no key of ours to hold and no card to add.

What it asks of you

It reaches a system of yours, so it needs your own credential rather than the publisher’s: GEMINI_API_KEY — Google Gemini API key from https://aistudio.google.com/apikey. You set it in your own client’s config, on your machine — this marketplace never holds it.

What it writes

One config entry your client uses to start the process locally. A local server runs with your file system and your network, which is why it is priced without metering.

What you approve

2 tools, with what each one reads, writes and reaches shown before you agree — the same list on every route above. Read the tool surface.

Tool surface

What the model actually sees. Descriptions are diffed on every release — see version history.

Only tools that write or reach the network

Read from its source code. A server that builds its tool list when it starts can list more tools than this.

Generate or edit images using Google Gemini.

Parameter	Type	Required	Notes
prompt	string	yes	Text description of the image to generate, or editing instruction when images are provided
images	array	no	File paths to input/reference images for editing. Omit for text-to-image generation. Per-model reference limits vary (gemini-3.1-flash-lite-image up to 14; others less) — the API enforces.
videos	array	no	File paths to input videos (mp4/mov/webm/etc, max 500MB each). The model watches the video and creates a NEW image from what it understood — thumbnails, posters, summary art. Not a frame grabber. gemi
model	string	no	Gemini image model ID. Defaults to the configured default (gemini-3.1-flash-lite-image). Validated at request time against the models your API key supports (discovered at startup). Common: gemini-3.1-
aspectRatio	string	no	Image aspect ratio (defers to the API — unsupported values are rejected by Gemini). Defaults to config value or 1:1. Current models support: 1:1, 2:3, 3:2, 3:4, 4:3, 4:5, 5:4, 9:16, 16:9, 21:9, plus 1
resolution	string	no	Image resolution. Defaults to config value or 1K. 512 only on gemini-3.1-flash-image; 2K/4K on gemini-3.1-flash-image and gemini-3-pro-image; gemini-3.1-flash-lite-image and gemini-2.5-flash-image are
outputDir	string	no	Directory to save the image. Defaults to config file outputDir, OUTPUT_DIR env var, or ~/gemini-images
filename	string	no	Base name for the saved file (e.g. 'hero-banner'). Extension added automatically. Duplicates get a version suffix (hero-banner-v2). Omit for auto-generated name.
subfolder	string	no	Subfolder within the output directory (e.g. 'landing-page'). Created automatically.
sessionId	string	no	Continue a multi-turn edit. Pass the sessionId from a previous response to refine that image across calls — the server keeps the prior turns as context.
seed	integer	no	Seed for reproducible generation. Same seed + prompt + model = same image.
useSearchGrounding	boolean	no	Legacy alias for grounding: 'web'. Prefer the grounding parameter.
grounding	string	no	Search grounding. 'web' = Google Search for real-world accuracy. 'web+image' adds image results (gemini-3.1-flash-image only; response includes searchEntryPointHtml which ToS requires displaying). Not
thinkingLevel	string	no	Thinking depth (gemini-3.1-flash family). Default MINIMAL = fast/cheap. Use HIGH for text-heavy or diagram/infographic renders.
removeBackground	object	no	Return a transparent PNG cutout in one call. Omit for a normal opaque image. Default mode 'auto' runs a local AI matte (no extra API cost; first use downloads a ~one-time model). Supplying color imp

Process an existing image locally using sharp. Crop, resize, remove background,

Parameter	Type	Required	Notes
imagePath	string	yes	Path to the image file to process
crop	object	no	Crop image. Use width+height for pixel-exact, or aspectRatio for ratio-based. Strategy controls where to crop from.
resize	object	no	Resize image. Maintains aspect ratio if only width or height given.
removeBackground	object	no	Remove background. mode 'auto' (AI matte, any subject), 'chroma' (green screen), or 'threshold' (white). Defaults: chroma if color set, else threshold.
trim	boolean	no	Auto-trim whitespace borders
format	string	no	Convert to format. Defaults to original format.
quality	integer	no	Output quality for JPEG/WebP (1-100). Default 90.
outputDir	string	no	Directory to save. Defaults to config file outputDir, OUTPUT_DIR env var, or ~/gemini-images
filename	string	no	Base name for saved file. Auto-versioned if duplicate.
subfolder	string	no	Subfolder within output directory

Free

$0 forever

Every tool, no call limit
No card, and mcprush charges nothing for it
Source published under MIT
Runs on your machine — nothing of it reaches our gateway
Nothing to cap, because mcprush bills nothing

Nothing here is billable: mcprush charges nothing to install Gemini Image or to call it, at any volume.

ATrust score 96 / 100

Release 0.6.2 · scanned 14 Oct 2026

Two independent axes, because powerful and malicious are different questions. The grade is threat only. The capability level is blast radius, and it is never a penalty on the grade — it is priced as one subtract-only term in the score, where you can see it.

Threat score100

Client score96

Findings2 threat · 0 capability

Capabilitymoderate

Coveragesource

Verificationrepo

Surface digest28beb4dc9bbc4df873075df58d510b07258c3786b0c930e39b1914fa4107e486

Score vectorthreat 100 → client 96

Term	Level	What it prices	Points
capability-exposure	moderate	capability blast radius (moderate) — client exposure if the model is manipulated	−3
verification-discount	repo	publisher verification (public source) — no provenance, but the source is public and inspectable	−1
coverage-honesty	source	inspection depth (source) — how much of the target the scan could see	−0

score = clamp(round(threatScore − E_cap − E_ver − E_cov), 0, 100). No model, no opinion, no network call at scoring time: the same methodology version against the same bytes always produces this number, and every line above is a field of the report rather than a summary of it.

What the scan could actually read

A grade is only as meaningful as its coverage, so the scanner publishes its own depth before it publishes its result.

Tool surfaceread

Implementation sourceread

Package metadataread

Live transportnot available

Tools were statically extracted from the published source (2 recovered), not enumerated from a running server. Tool-poisoning, Unicode-smuggling, capability and toxic-flow analysis ran on this inferred surface, but a mis-parsed registration could be missed or mis-attributed, so tool-derived findings are capped below “confirmed”. To grade the real runtime surface, scan the running server: --command "npx -y <package>".

Capability — what it could do if the model were manipulated

Tags derived from each tool’s schema and the implementation, not from what the tool calls itself. moderate is the level these add up to.

Tool	Capability tags	Why the tag was assigned
generate_image	no tags
process_image	no tags

Toxic-flow graph

The lethal trifecta, checked as a graph rather than as a checklist: untrusted input, a sensitive source and an external sink have to meet before there is a path worth worrying about.

The public result for this release does not print the flow graph, so there is nothing to show here. That is not the same as "no paths were found": what the scan did read is above, under coverage.

Supply chain and provenance

Integrityfirst-seen

Surface digest28beb4dc9bbc4df8

Scanned2026-10-14

This is the first scan of this surface here, so there is nothing yet to compare it against.

Every result on this tab comes from one deterministic pass over the published package — offline, rule by rule, and auditable line by line above. Same methodology version, same bytes, same score.