GPT Image MCP
by rayhankhilji
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| GPT_IMAGE_HOME | No | Directory for the profile, state, and debug captures. Default: ~/.gpt-image-mcp. | |
| GPT_IMAGE_DEBUG | No | Set to 1 for verbose logs on stderr. Default: 0. | 0 |
| GPT_IMAGE_MODEL | No | ?model= slug for new chats. Default: account default. | |
| GPT_IMAGE_WINDOW | No | Headed browser window mode: minimized, normal, or offscreen. Default: minimized. | minimized |
| GPT_IMAGE_CDP_URL | No | Attach to an existing Chrome instance instead of launching one, e.g. http://127.0.0.1:9222. | |
| GPT_IMAGE_SESSION | No | ChatGPT login/session value to seed an empty profile, such as the value printed by export-session --env or a Cookie-Editor export. Treat as a password. | |
| GPT_IMAGE_TIMEOUT | No | Per-job timeout, in seconds. Default: 600. | 600 |
| GPT_IMAGE_HEADLESS | No | Set to 1 for headless mode. Auto: on Linux without a screen, uses a virtual display Xvfb if installed, otherwise headless. | auto |
| GPT_IMAGE_HTTP_HOST | No | HTTP server bind host. | |
| GPT_IMAGE_HTTP_PORT | No | HTTP server bind port. | |
| GPT_IMAGE_WARM_TABS | No | Number of ready-to-type ChatGPT tabs the server keeps loaded so jobs start instantly. Default: 3. | 3 |
| GPT_IMAGE_API_GAP_MS | No | Starting gap between ChatGPT data requests, in milliseconds; adapts automatically. Default: 300. | 300 |
| GPT_IMAGE_HTTP_TOKEN | No | Bearer token required by the HTTP server and used by remote MCP clients. | |
| GPT_IMAGE_OUTPUT_DIR | No | Default folder where generated images are saved. Default: $GPT_IMAGE_HOME/images. | |
| GPT_IMAGE_PUBLIC_URL | No | Public base URL of the remote server. | |
| GPT_IMAGE_SERVER_URL | No | Make the CLI use a remote server instead of a local browser. | |
| GPT_IMAGE_PROFILE_DIR | No | Override the browser profile directory, allowing separate instances to use separate profiles. | |
| GPT_IMAGE_SEND_GAP_MS | No | Minimum gap between prompt submissions, in milliseconds. Default: 250. | 250 |
| GPT_IMAGE_SESSION_FILE | No | Path to a file containing a ChatGPT session/login to seed an empty profile. | |
| GPT_IMAGE_MAX_CALL_WAIT | No | Longest one tool call blocks, in seconds. Unfinished jobs are collected with wait_for_jobs. Default: 50. | 50 |
| GPT_IMAGE_MAX_IDLE_TABS | No | Number of finished chats kept open for instant edits. Default: 4. | 4 |
| GPT_IMAGE_MAX_PAGE_LOADS | No | Maximum number of ChatGPT pages loading at once. Default: 5. | 5 |
| GPT_IMAGE_BROWSER_CHANNEL | No | Browser channel: chrome, msedge, or chromium (Playwright's). Default: chrome. | chrome |
| GPT_IMAGE_EXECUTABLE_PATH | No | Explicit browser binary path. | |
| GPT_IMAGE_MAX_CONCURRENCY | No | Maximum number of chats generating at once; extra jobs queue. Default: 10. | 10 |
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
Server capabilities have not been inspected yet.
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
No tools | |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
This server cannot be deployed
Maintenance
ActivityMaintained
ResponsivenessNo recent community issues