hive.image-generation

1# Image generation
2 
3`image_generate` turns a text prompt into an image (and can edit existing
4images). It routes through the Hive image service to OpenAI's **gpt-image-2**;
5the cost is billed to the user's Hive credits exactly like an LLM call, so there
6is **no API key to configure**. Each generated image is also saved to disk.
7 
8## The call
9 
10```
11image_generate(
12    prompt: str,                       # required — what to draw
13    reference_images: list[str] = None,# local paths or http(s) URLs to edit/condition on
14    size: str = "1024x1024",           # 1024x1024 | 1536x1024 (landscape) | 1024x1536 (portrait) | auto
15    quality: str = "low",              # low only (medium & high disabled)
16    n: int = 1,                        # 1–4; each image is billed separately
17    output_format: str = "png",        # png | jpeg | webp
18    model: str = "gpt-image-2",
19)
20```
21 
22Defaults are deliberately cheap and fast. **`quality` is locked to `low`** —
23`medium` and `high` are disabled for cost control, and any request for a higher
24tier is automatically forced to `low`. Only raise `n` when the user explicitly
25wants variations.
26 
27### Writing the prompt
28Be concrete: name the subject, style (photo, flat vector, 3D, watercolor…),
29composition/framing, color palette, mood, and any **text to render** (gpt-image-2
30renders text well — quote it exactly, e.g. `the words "Launch Day" in bold`).
31 
32## Reference-image editing
33 
34Pass `reference_images` to edit, restyle, or compose from existing images —
35…