MiniMax H3 Image API

Autenticação

All requests carry the API key in the header:

Authorization: Bearer YOUR_API_KEY

Preços

An exact size is priced by area: up to 5,033,165 pixels at the 1K / 2K price, above that at the 4K price. The text prompt is free. Credit for n images plus every reference image is reserved when the task is created; on completion you are charged only for the images actually delivered and the rest is released. A request that fails entirely costs nothing.

Text-to-image

Async by default: create returns a taskId; poll GET ?task_id= until completed and read resultUrls (one URL per image), or receive a webhook via callback_url. In our tests a 1K 16:9 text-to-image request took about 10 seconds and returned 1360×768; a 1K poster on Highspeed took about 6 seconds. 4K and several images per request take longer.

# Text-to-image — create returns a taskId, then poll for the result.
curl -X POST https://api.apimodels.app/v1/images/generations \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "minimax-h3-image",
    "prompt": "A cozy cafe counter at golden hour, a ceramic latte cup with leaf latte art and rising steam, a croissant on a small plate, warm window light, shallow depth of field, film grain",
    "resolution": "1K",
    "aspect_ratio": "16:9",
    "n": 2
  }'
# → { "code": 200, "data": { "taskId": "...", "state": "pending", "model": "minimax-h3-image" } }

curl "https://api.apimodels.app/v1/images/generations?task_id=TASK_ID" \
  -H "Authorization: Bearer YOUR_API_KEY"
# → { "code": 200, "data": { "state": "completed", "resultUrls": ["https://r2.apimodels.app/...", "https://r2.apimodels.app/..."] } }

Editing · combining pictures

Include image (one) or images (up to 16) to edit or combine; refer to them as image 1, image 2 in the prompt. References can be public URLs or data URIs; anything above about 2.4 megapixels is scaled down before it is sent. An edit with one 1K reference image took about 16 seconds in our test.

# Edit / combine — 1 to 16 reference images (+$0.01 each, +$0.015 on Highspeed).
curl -X POST https://api.apimodels.app/v1/images/generations \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "minimax-h3-image",
    "prompt": "Replace the croissant on the plate with a slice of strawberry cheesecake. Keep the cup, saucer, latte art, steam, table, window light and everything else exactly the same.",
    "image": "https://example.com/cafe.jpg",
    "resolution": "1K",
    "aspect_ratio": "16:9"
  }'

# Several references — refer to them as image 1, image 2 in the prompt.
curl -X POST https://api.apimodels.app/v1/images/generations \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "minimax-h3-image",
    "prompt": "put the vase from image 2 on the table in image 1; keep the room, furniture and lighting unchanged",
    "images": ["https://example.com/dining-room.jpg", "https://example.com/vase.png"]
  }'

One call, OpenAI style

Pass a WIDTHxHEIGHT size and no callback_url and the request stays open until the images are ready; the reply is { created, data: [{ b64_json, url, width, height }], size }, and "response_format": "url" leaves out the base64 copy. The OpenAI SDK's images.generate gets this same response. Give the call a generous timeout: 4K or several images can take more than a minute.

# One call: an exact WIDTHxHEIGHT size and no callback_url — the images come back in this response.
curl -X POST https://api.apimodels.app/v1/images/generations \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "minimax-h3-image-highspeed",
    "prompt": "A minimalist jazz night poster: a deep blue background, a gold trumpet silhouette, the headline \"BLUE NOTE FRIDAY\" in large elegant serif letters, and a smaller line \"Live at The Lantern Room - 9 PM\" at the bottom",
    "size": "2720x1536",
    "response_format": "url"
  }'
# → { "created": ..., "data": [{ "url": "https://r2.apimodels.app/...", "width": 2720, "height": 1536 }], "size": "2720x1536" }

Parâmetros

Limits

• Masks (mask / mask_url), transparent backgrounds (background: "transparent"), negative_prompt and seed are not supported and return 400.
• n above 10, more than 16 reference images, a shape beyond 1:3–3:1 and a size outside 655,360–20,000,000 pixels all return 400.
• Edits always use prompt enhancement; disable_prompt_enhanced applies to text-to-image only.
• There is no seed, so the same request returns a different picture each time.
• Parameters that cannot be honoured return 400 with the reason instead of being silently ignored.

Not every request succeeds

Content review can refuse a prompt or an image, and the upstream occasionally returns a temporary error. When you ask for several images you pay only for the ones delivered; if the whole request fails nothing is charged and the reserved credit is released. The response carries the failure reason: for review refusals, change the prompt or image; for temporary errors, retry. Result files are deleted after 30 days, so download or re-host them promptly.

Try it in the playground: MiniMax H3 Image / Highspeed (one image per run there; use the API for several images, exact sizes and PNG / WebP). For transparent assets, see Seedream 5.5.