✨ GPT Image 2.6 is now available. Come and experience it!

MiniMax-H3

Hotminimax/MiniMax-H3
Capabilities
4
Output resolution
2K
768p2K
Generation duration
4–15s
4s5s6s7s
Release date
2026-07-31
MiniMax

Playground

Input

Output

Pricing

Pricing breakdown
Usage-based pricing, broken down by billing item at your current rates.
15% off

Input images after the fifth

Standard rate
0.04 credits0.034 credits

Input video duration (s)

2 price tiers

Output resolution: 768p
0.08 credits0.068 credits
Output resolution: 2K
0.13 credits0.1105 credits

Video duration (s)

2 price tiers

Output resolution: 768p
0.08 credits0.068 credits
Output resolution: 2K
0.13 credits0.1105 credits

Billing items are additive. Only one matching price tier applies to each item. Each request is calculated exactly from actual usage, and the total is rounded up once to the nearest 0.000001 credits.

API

cURL example
Request type
POST/v2/video_generation
AsynchronousNative
CREATE_RESPONSE=$(curl "https://api.modelrover.ai/v2/video_generation" \  -X POST \  -H "Authorization: Bearer $API_KEY" \  -H "Content-Type: application/json" \  -d '{  "model": "minimax/MiniMax-H3",  "content": [    {      "type": "text",      "text": "Explain model gateways in one sentence."    }  ],  "resolution": "768p",  "duration": 5,  "ratio": "16:9"}')# Requires jq.TASK_ID=$(printf '%s' "$CREATE_RESPONSE" | jq -r '.task_id')for attempt in $(seq 1 120); do  RESPONSE=$(curl -sS "https://api.modelrover.ai/v1/tasks/"$TASK_ID -H "Authorization: Bearer $API_KEY")  STATUS=$(printf '%s' "$RESPONSE" | jq -r '.task.status')  case "$STATUS" in queued|running) sleep 10 ;; *) printf '%s' "$RESPONSE"; break ;; esacdone

Set your API_KEY environment variable before running. Keep your key server-side.

API documentation

Uptime