Image Generation

1 from meshapi import ImageGenerationParams
2 
3 result = client.images.generate(
4     ImageGenerationParams(
5         model="openai/gpt-image-1",
6         prompt="A watercolor of a fox in a snowy forest",
7         n=1,
8         size="1024x1024",
9         quality="high",
10         output_format="webp",
11     )
12 )
13 
14 print(result.data[0].url)

Where the image lands depends on the model and response_format. With the default (or response_format="url"), some models — including openai/gpt-image-1 — return the image inline as a data: URI in url rather than in b64_json. Use result.data[0].image_bytes() to get raw bytes regardless of shape (it decodes b64_json or a data: URL), or pass response_format="b64_json" for predictable base64. A remote http(s) url raises — fetch it yourself.

1 with open("out.png", "wb") as f:
2     f.write(result.data[0].image_bytes())

Streaming

1 for chunk in client.images.stream(
2     ImageGenerationParams(
3         model="openai/gpt-image-1",
4         prompt="A watercolor of a fox in a snowy forest",
5         n=1,
6         size="1024x1024",
7         quality="high",
8         output_format="webp",
9     )
10 ):
11     if chunk.status == "processing":
12         print("Generating...")
13     elif chunk.data:
14         print("Done:", chunk.data[0].url)

Async

1 from meshapi import AsyncMeshAPI, ImageGenerationParams
2 
3 async_client = AsyncMeshAPI()
4 
5 async for chunk in async_client.images.stream(
6     ImageGenerationParams(
7         model="openai/gpt-image-1",
8         prompt="A watercolor of a fox in a snowy forest",
9         n=1,
10         size="1024x1024",
11         quality="high",
12         output_format="webp",
13     )
14 ):
15     if chunk.status == "processing":
16         print("Generating...")
17     elif chunk.data:
18         print("Done:", chunk.data[0].url)

Parameters

Field	Type	Notes
`prompt`	`str`	Required
`model`	`str \| None`	e.g. `"openai/gpt-image-1"`
`n`	`int \| None`	Number of images (1–10)
`size`	`str \| None`	`"auto"`, `"1024x1024"`, `"WIDTHxHEIGHT"`
`quality`	`str \| None`	`"auto"`, `"low"`, `"medium"`, `"high"`, `"hd"`, `"standard"`
`response_format`	`"url" \| "b64_json" \| None`
`output_format`	`"png" \| "jpeg" \| "webp" \| None`	OpenAI only
`stream`	`bool \| None`	Enable SSE keep-alive streaming