Claude visitors: new accounts receive $1.00 in bonus API credit within 30 days. One per person; usage credits only.
Use GPT Image in WorkBuddy.
GPT Image 2 is OpenAI's dedicated image generation model — text-to-image and instruction-based image editing with selectable aspect ratios and up to 4K output resolution. WorkBuddy calls it as an MCP tool on the same RunAPI server and RUNAPI_API_KEY used for chat, with nothing extra to install.
Use the runapi MCP server to generate an image with OpenAI GPT Image 2.
Requirements:
- Call the runapi create_task tool with the text_to_image action.
- Set the model to "gpt-image-2".
- Write a descriptive prompt. GPT Image 2 follows natural language instructions
closely — describe layout, style, and text overlays.
- Optionally set output_resolution to 1k, 2k, or 4k. Default is 1k.
- The task is async. Poll it with get_task until it completes, then report the
output image URL.
{
"mcp_servers": {
"runapi": {
"url": "https://runapi.ai/mcp",
"transport": "streamable_http",
"headers": {
"Authorization": "Bearer $RUNAPI_API_KEY"
}
}
}
}
{
"task_id": "tsk_abc123",
"status": "completed",
"model": "gpt-image-2",
"output": {
"image_url": "https://media.runapi.ai/outputs/tsk_abc123.png",
"output_resolution": "2k"
}
}
Use GPT Image in WorkBuddy in three steps
Add the RunAPI MCP server
Register an MCP server named runapi in WorkBuddy, pointing at https://runapi.ai/mcp over streamable HTTP with your key read from RUNAPI_API_KEY. If RunAPI is already connected for chat, GPT Image tools are exposed by the same server — no additional setup needed.
url: https://runapi.ai/mcp
auth: Bearer $RUNAPI_API_KEY
Ask WorkBuddy for a GPT Image task
Prompt the workbench agent to call the runapi create_task tool with the text_to_image action and model gpt-image-2, including a descriptive prompt with layout and style instructions. Set output_resolution to 2k or 4k for higher detail. For editing existing images, have it use the edit_image action with model gpt-image-2 and source_image_urls.
create_task(service="gpt-image-2",
action="text_to_image",
model="gpt-image-2",
prompt="...",
output_resolution="2k")
Get the result
The tool call returns a task_id immediately. WorkBuddy polls with the get_task tool until the status changes to completed, then reports the output image URL in the conversation.
get_task(task_id="tsk_abc123")
GPT Image task parameters
| Parameter | Type | Description |
|---|---|---|
model |
string |
gpt-image-2. The text_to_image and edit_image actions use the same model ID. |
prompt |
string |
Required. Natural language description of the desired image. Supports detailed instructions for layout, text overlays, and style. |
output_resolution |
string |
Optional. Output resolution — 1k (default), 2k, or 4k. For text_to_image, higher resolution costs more per image; edit_image bills one rate at every resolution. |
aspect_ratio |
string |
Optional. Defaults to auto. Supports 1:1, 3:2, 2:3, 4:3, 3:4, 16:9, 9:16, and more. |
source_image_urls |
array |
Required for the edit_image action. One or more URLs of source images to edit. |
What is GPT Image on WorkBuddy?
GPT Image 2 is OpenAI's dedicated image model that works more like a structured design assistant than a keyword-driven generator. You give it a production brief — layout, text placement, style constraints — and it follows the instructions closely. Users report strong character consistency, clean text rendering, and readable layouts, though quality can degrade through multiple iterative refinement passes. WorkBuddy agents call it as a tool on the same RunAPI MCP server used for chat.
GPT Image use cases
UI mockups and app design wireframes
Generate detailed UI mockups from written descriptions, including button layouts, navigation bars, and placeholder content that reads correctly.
Text-heavy posters and presentations
Create exact-text posters, infographics, and presentation slides where the text content is specified in the prompt and rendered with clean, readable layouts.
Character design sheets
Generate character design sheets with turnaround views from a single description, maintaining consistent features across multiple angles.
GPT Image + WorkBuddy questions
Yes. WorkBuddy speaks MCP natively, so once RunAPI is registered as an MCP server, GPT Image 2 is available as a tool. Ask the workbench agent for a text_to_image task with model gpt-image-2 — the same RUNAPI_API_KEY used for chat covers it, with no extra skills or plugins.
GPT Image 2 is OpenAI's dedicated image generation model with higher quality and up to 4K output. GPT-4o Image generates images within a chat context but is limited to 1:1, 3:2, or 2:3 aspect ratios. Both are available through RunAPI — use gpt-image-2 for standalone generation and gpt-4o-image for chat-integrated image output.
Set aspect_ratio to auto or a ratio from 1:1 to 21:9 or 9:21, and output_resolution to 1k, 2k, or 4k. Resolution defaults to 1k.
GPT Image 2 is billed per image. For text_to_image, the rate depends on output resolution: 1k is the lowest cost, 2k is mid-range, and 4k is the most expensive. edit_image bills the same per-image rate at every resolution. Check the RunAPI pricing page for current per-image rates. Failed generations are not charged.
Yes. Ask WorkBuddy to run an edit_image task with model gpt-image-2. Pass the source image URLs in source_image_urls and describe the edit in the prompt — for example, "change the background to a beach sunset" or "add a red hat to the person." GPT Image 2 follows natural language editing instructions.
WorkBuddy general setup
Not configured yet? Start with the RunAPI MCP setup guide for WorkBuddy.
WorkBuddy setup guide →Try GPT Image in WorkBuddy today.
Get a free RunAPI key, add the MCP server to WorkBuddy, and start generating and editing images with OpenAI GPT Image 2 from your workbench.