Inference Logoinference shell

MiniMax H3 Image

MiniMax H3 Image. Text-to-image and editing with up to 16 reference images. 1K to 4K output, any aspect ratio from 1:3 to 3:1, up to 10 images per run, built-in prompt enhancement.

text to imageimage editing
run in browser →run via APIrun with your agent
maker
minimax
category
image
input → output
text → image
runs via
api, javascript and python sdk, belt cli, mcp
updated
2026-11-10
price
$2.30/M text in, $1.50/M image in, $3.80/M image out (token-based)

api reference

about

minimax h3 image. text-to-image and editing with up to 16 reference images. 1k to 4k output, any aspect ratio from 1:3 to 3:1, up to 10 images per run, built-in prompt enhancement.

get an api key

create one in settings → api keys. the examples read it from INFERENCE_API_KEY.

bash
1export INFERENCE_API_KEY="inf_your_key"

1. calling the api

install the client

nothing to install. any http client works; the examples use curl, and jq to read responses. every response wraps its result in a data field.

run and get result

send "wait": true to hold the connection until the task finishes. without it the call returns right away with the task still running.

bash
12# "wait": true holds the connection until the task finishes.3# the response is {"data": {"id", "status", "status_text", "output", ...}}4curl -X POST https://api.inference.sh/apps/run \5  -H "Authorization: Bearer $INFERENCE_API_KEY" \6  -H "Content-Type: application/json" \7  -d '{8    "app": "minimax/h3-image",9    "input": {10      "prompt": "A ceramic teapot shaped like a lighthouse on a weathered wooden table, soft morning window light, shallow depth of field, product photography.",11      "aspect_ratio": "16:9"12    },13    "wait": true14  }'

stream live updates

get real-time progress updates as the task runs. ideal for showing progress bars, partial results, or long-running tasks. send "stream": true and the response is a server-sent event stream that ends when the task does.

bash
12# "stream": true answers with server-sent events until the task finishes3curl -N -X POST https://api.inference.sh/apps/run \4  -H "Authorization: Bearer $INFERENCE_API_KEY" \5  -H "Content-Type: application/json" \6  -d '{7    "app": "minimax/h3-image",8    "input": {9      "prompt": "A ceramic teapot shaped like a lighthouse on a weathered wooden table, soft morning window light, shallow depth of field, product photography.",10      "aspect_ratio": "16:9"11    },12    "stream": true13  }'

run in the background

start the task and get its id back right away, then check on it later. for long tasks, or when you'd rather be called back, see webhooks.

bash
1# without "wait" the call returns the task right away (status 1, "received")2TASK_ID=$(curl -s -X POST https://api.inference.sh/apps/run \3  -H "Authorization: Bearer $INFERENCE_API_KEY" \4  -H "Content-Type: application/json" \5  -d '{6    "app": "minimax/h3-image",7    "input": {8      "prompt": "A ceramic teapot shaped like a lighthouse on a weathered wooden table, soft morning window light, shallow depth of field, product photography.",9      "aspect_ratio": "16:9"10    }11  }' | jq -r '.data.id')1213# later: read its state. done when status is 10 (completed), 11 (failed) or 12 (cancelled)14curl -s https://api.inference.sh/tasks/$TASK_ID \15  -H "Authorization: Bearer $INFERENCE_API_KEY" | jq '.data | {status, status_text, output}'

2. authentication

the api uses api keys for authentication. see the authentication docs for detailed setup instructions.

3. files

file inputs are automatically handled by the sdk. you can pass local paths, urls, or base64 data.

automatic upload

pass urls directly or use data uris for inline content.

bash
1# pass URLs or data URIs as input values2curl -X POST https://api.inference.sh/apps/run \3  -H "Authorization: Bearer $INFERENCE_API_KEY" \4  -H "Content-Type: application/json" \5  -d '{6    "app": "minimax/h3-image",7    "input": {8      "image": "https://example.com/image.png",9      "audio": "data:audio/mp3;base64,..."10    },11    "wait": true12  }'

manual upload

you can also upload files manually and use the returned url.

bash
1# 1. create the file: the response has an upload_url and the file's uri2FILE=$(curl -s -X POST https://api.inference.sh/files \3  -H "Authorization: Bearer $INFERENCE_API_KEY" \4  -H "Content-Type: application/json" \5  -d '{"files": [{"filename": "image.png", "content_type": "image/png"}]}')67# 2. PUT the bytes to upload_url8curl -X PUT "$(echo "$FILE" | jq -r '.data[0].upload_url')" \9  -H "Content-Type: image/png" \10  --data-binary @image.png1112# 3. pass the uri as an input value13echo "$FILE" | jq -r '.data[0].uri'

4. webhooks

get notified when a task completes by providing a webhook url. when the task reaches a terminal state (completed, failed, or cancelled), a POST request is sent to your url with the task result.

bash
1curl -X POST https://api.inference.sh/apps/run \2  -H "Authorization: Bearer $INFERENCE_API_KEY" \3  -H "Content-Type: application/json" \4  -d '{5    "app": "minimax/h3-image",6    "input": {7        "prompt": "A ceramic teapot shaped like a lighthouse on a weathered wooden table, soft morning window light, shallow depth of field, product photography.",8        "aspect_ratio": "16:9"9    },10    "webhook": "https://your-server.com/webhook"11  }'

webhook payload

your endpoint receives a JSON POST with the task result:

json
1{2  "event": "task.completed",3  "timestamp": "2024-01-15T10:30:05Z",4  "data": {5    "id": "task_abc123",6    "short_id": "abc123",7    "status": 10,8    "status_text": "completed",9    "output": { ... },10    "created_at": "2024-01-15T10:30:00Z",11    "updated_at": "2024-01-15T10:30:05Z"12  }13}
eventstringtask.completed, task.failed or task.cancelled
timestampstringiso timestamp of the delivery
data.idstringtask id
data.short_idstringshort task id
data.statusnumberterminal status (10=completed, 11=failed, 12=cancelled, 13=rejected)
data.status_textstringthe status as a word: completed, failed or cancelled
data.outputobjecttask output (when completed)
data.errorstringerror message (when failed, omitted otherwise)
data.session_idstringsession id (when the task ran in a session, omitted otherwise)
data.created_atstringiso timestamp
data.updated_atstringiso timestamp

5. schema

input

promptstring*

what to generate, or how to change the reference images. refer to references as picture 1, picture 2, in upload order. up to 32000 characters.

example: "A ceramic teapot shaped like a lighthouse on a weathered wooden table, soft morning window light, shallow depth of field, product photography."minLength:1maxLength:32000
pattern: \S
imagesarray

reference images for editing or composition, up to 16. jpg, png, webp, heic or heif; each side 128 to 8192 px; long side at most 4x the short side. leave empty for text-to-image.

aspect_ratiostring

output aspect ratio as x:y (e.g. 16:9, 3:4, 21:9), between 1:3 and 3:1, or auto to let the model choose from the prompt and references.

default: "auto"example: "16:9"
pattern: ^(auto|[1-9][0-9]{0,2}:[1-9][0-9]{0,2})$
resolutionstring

output area: 1k (about 1024x1024), 2k (about 2048x2048) or 4k (about 4096x4096). width and height follow the aspect ratio, e.g. 2k at 16:9 is 2720x1536.

default: "2K"
options:"1K""2K""4K"
sizestring

exact output size as widthxheight, e.g. 2720x1536. overrides aspect_ratio and resolution. area 655,360 to 20,000,000 px, ratio between 1:3 and 3:1.

pattern: ^[1-9][0-9]{2,3}x[1-9][0-9]{2,3}$
num_imagesinteger

number of images to generate.

default: 1min:1max:10
enhance_promptboolean

let the model expand the prompt before generating. always on when reference images are given. when off with aspect_ratio auto, the image is 1:1.

default: true
output_formatstring

image file format.

default: "jpeg"
options:"jpeg""png""webp"

output

imagesarray*

the generated images.

revised_promptsarray

the prompt the model used for each image, after enhancement.

other image apps

GPT Image 2.6 Flare

openai/gpt-image-2-6-flare

$30.00/M output, $5.00/M text input, $8.00/M image input, $0.10/M SVG input, $0.50/M SVG output

GPT Image 2.6 Sunburst

openai/gpt-image-2-6-sunburst

$30.00/M out, $5.00/M text in, $8.00/M img in; SVG $0.10/$0.50/M in/out

Gemini Nano Banana 2.1

google/gemini-nano-banana-2-1

$1.50/M input, $0.0336–$0.1134/image (1K–4K), +$0.014/query Google Search

Grok Imagine Image 2.0

xai/grok-imagine-image-2-0

$0.04/img low 1k, $0.06/img low 2k, $0.06/img med 1k, $0.08/img med 2k; $0.01/source img (edits)

GPT Image 2.5 Sunburst

openai/gpt-image-2-5-sunburst

$0.006-$0.21 per 1024x1024 image: low $0.006, medium $0.013, high $0.053, xhigh $0.094, max $0.21. Larger sizes cost more; edits add ~$0.008 per reference image.

more from minimax

MiniMax Music Cover

minimax/music-cover

$0.15/cover

MiniMax Speech 2.8 HD

minimax/speech-2-8-hd

$0.0001/char

MiniMax Music 3.0

minimax/music-3-0

$0.15/track

MiniMax Speech 2.8 Turbo

minimax/speech-2-8-turbo

$60.00/M chars

new models

  1. Pika H3 Flex (self-hosted)

    pika/pika-h3-flex-oss

    videoGPU time, billed per millisecond

  2. Pika H3 Flex Audio Stems

    pika/pika-h3-flex-audio-stems

    audio$0.001/stem-sec (3 stems: dialogue, music, effect)

  3. Pika H3 Flex Upscale

    pika/pika-h3-flex-upscale

    video$0.05-$0.14/sec (480p src), $0.03-$0.12/sec (768p src); 1080p/2K/4K targets

  4. Pika H3 Flex Reference to Video

    pika/pika-h3-flex-reference

    video$0.06/sec 480p, $0.12/sec 768p, $0.15/sec 1080p, $0.16/sec 2k, $0.20/sec 4k (output+input video sec). Stems: $0.001/sec

  5. Pika H3 Flex

    pika/pika-h3-flex

    video$0.02/sec 480p, $0.04/sec 768p, $0.07/sec 1080p, $0.08/sec 2K, $0.16/sec 4K; stems +$0.001/stem-sec

ready to run MiniMax H3 Image?

try in browserbrowse all apps

we use cookies

we use cookies to ensure you get the best experience on our website. for more information on how we use cookies, please see our cookie policy.

by clicking "accept", you agree to our use of cookies.
learn more.