get_savings_stats
Retrieve cumulative token and bandwidth savings achieved through VisionSqueezer optimizations.
LABSCO SUMMARY
A preprocessing step between your images and the vision bill.
Tile alignment is the mechanism: vision APIs charge by tile, and an image that does not land on those boundaries pays for padding it does not use. sandbox_execute is the more aggressive option — crop and binarize down to the region that carries the answer, and the rest never becomes tokens at all.
What it is
An image preprocessing MCP server for vision models. It resizes to tile boundaries, strips padding and re-encodes so an image costs fewer tokens to look at, and it works with any agent that speaks MCP rather than a single provider.
What you get
- optimize_image — snap an image's dimensions to tile boundaries, remove padding and re-encode it for a vision API
- optimize_image_batch — up to 64 images in one call, each taking the same arguments; a failure returns an error entry for that image rather than sinking the batch
- sandbox_execute — run a sequence of atomic operations (crop, grayscale, binarize, resize, contrast, brightness) to extract only the region that matters
- get_savings_stats — the cumulative token and bandwidth savings from optimizations so far
Requirements
Nothing to configure — clients run npx -y vision-squeezer over stdio, or npx vision-squeezer install to pick the client, method and scope interactively. In Claude Code the plugin path bundles the server together with the /vision-stats, /vision-doctor and /vision-upgrade skills. There is also a Rust crate and a Python package of the same name.
Setup effort
One command — npx vision-squeezer install
USE CASES
01
Feeding a screenshot to a vision model
Snap it to tile boundaries and drop the padding before the model looks at it.
02
Processing a folder of images
Optimize up to 64 in one call and read which ones failed.
03
Reading one panel of a large diagram
Crop to the region, raise contrast, and binarize so only that part becomes tokens.
04
Showing what the preprocessing bought
Read the cumulative token and bandwidth savings.
4 tools
Counted once with only startup setup. Other versions, modes or optional credentials may expose a different list.
optimize_image
Resize and optimize an image for LLM vision APIs. Snaps dimensions to tile boundaries, removes padding, and re-encodes to minimize token consumption.
optimize_image_batch
Optimize multiple images in one call (max 64). Each entry takes the same arguments as optimize_image. Returns a per-image results array; a failing image yields an error entry without failing the whole batch.
sandbox_execute
Think in Code: Execute a sequence of atomic operations (crop, grayscale, binarize, resize, contrast, brightness) on an image to extract only necessary context and slash tokens.
BEFORE YOU INSTALL
What to know first
Tools you get
4 tools when we connected.
Requires
npx on PATH
CONTEXT COST
Light Standing cost
Rated by the size of the 4 tool definitions that ride along on every turn, not by how many there are
INSTALLATION
Connect your client
Server setup
TERMINAL
npx vision-squeezer installMCP Clients
Claude Code — plugin
CLAUDE CODE — PLUGIN · IN CLIENT
/plugin marketplace add eralpozcan/vision-squeezer
/plugin install vision-squeezer-mcp@vision-squeezerClaude Code — npx
CLAUDE CODE — NPX · TERMINAL
claude mcp add --scope user vision-squeezer -- npx -y vision-squeezerClaude Desktop, Windsurf, Gemini CLI, Kiro, Antigravity
JSON
{
"mcpServers": {
"vision-squeezer": {
"command": "npx",
"args": [
"-y",
"vision-squeezer"
]
}
}
}Cursor
CURSOR · JSON
{
"mcpServers": {
"vision-squeezer": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"vision-squeezer"
]
}
}
}VS Code
VS CODE · JSON
{
"servers": {
"vision-squeezer": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"vision-squeezer"
]
}
}
}JetBrains
JETBRAINS · JSON
{
"servers": {
"vision-squeezer": {
"command": "npx",
"args": [
"-y",
"vision-squeezer"
]
}
}
}Codex
CODEX · TOML
[mcp_servers.vision-squeezer]
command = "npx"
args = ["-y", "vision-squeezer"]Qwen Code
QWEN CODE · TERMINAL
qwen mcp add vision-squeezer -- npx -y vision-squeezerOpenCode
OPENCODE · JSON
{
"mcp": {
"vision-squeezer": {
"type": "local",
"command": [
"npx",
"-y",
"vision-squeezer"
],
"enabled": true
}
}
}Kimi CLI
KIMI CLI · TERMINAL
kimi mcp add vision-squeezer -- npx -y vision-squeezerZed
ZED · JSON
{
"context_servers": {
"vision-squeezer": {
"command": "npx",
"args": [
"-y",
"vision-squeezer"
]
}
}
}
