Skip to main content
Glama
iconben

Z-Image Studio

by iconben
OverviewSchemaSourceRelated ServersDiscussionsAdmin
Python
Hybrid

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault
HOSTNoThe host to bind the server to when running in 'serve' mode.0.0.0.0
PORTNoThe port to bind the server to when running in 'serve' mode.8000
ZIMAGE_BASE_URLNoBase URL for generated links. Used to construct absolute URLs for ResourceLinks when using SSE or Streamable HTTP transports.
ZIMAGE_DISABLE_MCPNoSet to '1' to disable all MCP endpoints (default is '0').0
Z_IMAGE_STUDIO_DATA_DIRNoOverride the default data directory location where the database and LoRAs are stored.
HSA_OVERRIDE_GFX_VERSIONNoFor AMD ROCm users on Linux, manual override for gfx version (e.g., '10.3.0' for RDNA2, '11.0.0' for RDNA3) if the GPU is not officially supported.
Z_IMAGE_STUDIO_OUTPUT_DIRNoOverride the default output directory location where generated images are saved.
ZIMAGE_ENABLE_TORCH_COMPILENoForce enable torch.compile optimization (experimental). Set to '1' to enable. By default disabled for Python 3.12+ due to known compatibility issues.

Instructions

Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.

This server publishes no instructions, or was last inspected before Glama recorded them.

Capabilities

Features and capabilities supported by this server

Protocol revision2025-11-25

CapabilityDetails
tools
{
  "listChanged": false
}
prompts
{
  "listChanged": false
}
resources
{
  "subscribe": false,
  "listChanged": false
}
experimental
{}

Tools

Functions exposed to the LLM to take actions

NameDescription
generateA
Generate an image from a text prompt.

Returns a consistent content array for both stdio and SSE transports:
1. TextContent: Enhanced metadata including generation info and file details
2. ResourceLink: Main image file reference with context-appropriate URI:
   - SSE: Absolute URL built from request context (X-Forwarded-* headers), ZIMAGE_BASE_URL, or relative path
   - Stdio: file:// URI for local access
3. ImageContent: Thumbnail preview (base64 PNG, max 400px)

URI Building Priority (SSE):
1. Context parameter (ctx.request_context.request) - builds absolute URL from request headers
2. ZIMAGE_BASE_URL environment variable - uses configured base URL
3. Relative URL - fallback when no other method available

File metadata (filename, file_path) is in TextContent to avoid duplication in ResourceLink.

For long-running operations (high steps/large images), this function will:
- Send progress notifications at key milestones via ctx.report_progress()
- Handle client disconnections gracefully
list_modelsA

List available image generation models and hardware recommendations.

list_historyB

List recent image generations history.

Returns no records when history is disabled (disable_history config or
ZIMAGE_DISABLE_HISTORY env var), even if the database holds older entries.

Prompts

Interactive templates invoked by user choice

NameDescription

No prompts

Resources

Contextual data attached and managed by the client

NameDescription

No resources

Deploy Server
Shareable, User-Owned Context BackendPeople, applications, and agents read, write, and coordinate together in one context lake. Reach it over MCP, a CLI, a Python SDK, or REST.fulcradynamics.comAd

TDQS

A3.6/5.0

Scored across 3 tools

Disambiguation5/5

Each tool has a clearly distinct target: list_history reads past generations, list_models reads available models/hardware, and generate creates a new image. The two list_* tools differ in resource domain (history vs. models), so misselection is unlikely.

Naming Consistency4/5

Two tools follow a clean verb_noun pattern (list_history, list_models), while generate is a bare verb without a noun object. This is a minor deviation but still readable and predictable overall.

Tool Count4/5

Three tools is lean but each earns its place: generation, model discovery, and history review cover the essential loop of a focused image studio. It sits at the low end, but nothing feels redundant or missing from the core scope.

Completeness3/5

The core workflow (discover models → generate → review history) is covered, and images are returned via ResourceLink/ImageContent. However, there is no way to fetch a specific history entry, delete or prune history, or check status of a long-running generation, leaving notable gaps.

Maintenance

ActivityActive
ResponsivenessNo issues

Resources