Image editing for image_edit_oai - gpt-image-1 tool

[mansiibm]

Hi, I have enabled the gpt-image-1 tool for my instance and using - v0.8.0-rc2. I am using the model from azure and able to generate the image but not edit it.

On the playground when I try to edit it works but in the application it fails.

and on the playground-

The image can be edited on the playground but not in the application.

I tried setting up an agent using gpt-4.1 from azure openAI and use the tool.

Logs-

2025-08-13T09:19:46.026358284Z {"data":{"error":{"code":"404","message":"Resource not found"}},"headers":{"apim-request-id":"c4cfc8ea-f5ca-4f8e-abfe-b449e53b67f2","content-length":"56","content-type":"application/json","date":"Wed, 13 Aug 2025 09:19:45 GMT","strict-transport-security":"max-age=31536000; includeSubDomains; preload","x-content-type-options":"nosniff"},"level":"error","message":"[image_edit_oai] Problem editing the image: The server responded with status 404: Request failed with status code 404","stack":"AxiosError: Request failed with status code 404\n at settle (/app/node_modules/axios/dist/node/axios.cjs:2031:12)\n at IncomingMessage.handleStreamEnd (/app/node_modules/axios/dist/node/axios.cjs:3148:11)\n at IncomingMessage.emit (node:events:536:3...","status":404,"timestamp":"2025-08-13T09:19:46.026Z"}

Even if I use 4.1 from openai it fails with 404.

[danny-avila]

I am also using gpt-image-1 via Azure.

# AZURE GPT-IMAGE-1
IMAGE_GEN_OAI_API_KEY=your-key
IMAGE_GEN_OAI_BASEURL=https://my-id-westus3.cognitiveservices.azure.com/openai/deployments/gpt-image-1/
IMAGE_GEN_OAI_AZURE_API_VERSION=2025-04-01-preview

[mansiibm]

Seems like you have deployed this model as chat model from azure OpenAI. The config I shared with you - will only add it to tools. In my yaml I just have chat completion models.

[jsiegmund]

I guess that's the part where I go wrong then, but I am not following on where to add that config then. So your gpt-image-1 model config is NOT in your librechat.yaml? Where is it then?

[mansiibm]

I have deployed the model in the azure open AI resource. That's it. After that i set these variables in the env file-

AZURE GPT-IMAGE-1

IMAGE_GEN_OAI_API_KEY=your-key
IMAGE_GEN_OAI_BASEURL=the url
IMAGE_GEN_OAI_AZURE_API_VERSION=2025-04-01-preview

and then I just add the tool name in the included tools section I have -

includedTools:
["youtube", "google", "calculator", "azure-ai-search", "image_gen_oai"]

and 'tools' in the capabilities for agents -

agents:
capabilities:
[
"file_search",
"actions",
"tools",
"artifacts",
"chain",
"execute_code",
"context",
"ocr",
]

That's all. I do not add the deployment the model in my deployment section for azure in my yaml.

The above is the only config I need to run the gpt-image-1 model.

In my yaml , my azure config is like this -

endpoints: azureOpenAI: groups: - group: "${RESOURCE_GROUP_NAME}" apiKey: "${AZURE_OPENAI_KEY}" instanceName: "${AZURE_OPENAI_RESOURCE_NAME}" # Mark this group as assistants compatible # version must be "2024-02-15-preview" or later version: "2025-01-01-preview" models: gpt-4.1: deploymentName: gpt-4.1 gpt-4.1: true gpt-4.1-mini: deploymentName: gpt-4.1-mini gpt-4.1-mini: true o4-mini: deploymentName: o4-mini o4-mini: true o3: deploymentName: o3 o3: true gpt-5: deploymentName: gpt-5 gpt-5: true gpt-5-mini: deploymentName: gpt-5-mini gpt-5-mini: true gpt-5-nano: deploymentName: gpt-5-nano gpt-5-nano: true ,

That's it , no mention of the image model.

[jsiegmund]

Great! I managed to get it working with your help. It's quite easy to mess up the configuration it seems, but it works now. Thanks a lot!

[mansiibm]

Glad it worked for you !!!