Hailuo API (MiniMax video) — models, credit pricing, and the live alternative

Hailuo (海螺) is MiniMax's video generation family — Hailuo 02 Standard, 02 Pro and 2.3 Pro — known for smooth, physically plausible motion and particularly strong image-to-video. The Hailuo API (often searched as the MiniMax video API) generates short clips programmatically. This guide explains the family, its credit-based pricing, and how to run text-to-video on Kunavo today.

The Hailuo model family

These rates are reserved in the catalog and go live once MiniMax's USD pricing stabilizes — they are not billable today.

Hailuo API pricing

MiniMax bills Hailuo in credits, so the effective USD cost depends on which credit bundle you buy and which reseller you use — third-party quotes vary widely and aren't directly comparable. Kunavo publishes transparent per-clip USD pricing for its live video model instead of a credit abstraction:

Text-to-video on Kunavo today

Kunavo exposes video generation through one OpenAI-style endpoint, POST /v1/video/generations. The live model is Veo 3 — cinematic clips with native audio, three price tiers. The request shape is identical to what a Hailuo call would be — only the model string differs:

import requests

# Live text-to-video on the same OpenAI-style endpoint. Swap the model
# string to Hailuo once it's enabled — the request shape is identical.
resp = requests.post(
    "https://api.kunavo.com/v1/video/generations",
    headers={"Authorization": f"Bearer {API_KEY}"},
    json={
        "model": "veo-3",   # live today; Hailuo on the roadmap
        "prompt": "a surfer carving a wave at golden hour, drone shot",
        "duration": 5,
        "aspect_ratio": "16:9",
    },
    timeout=600,
)
print(resp.json()["data"][0]["url"])

The live alternative: Veo 3

Google Veo 3 is the strongest generally available text-to-video model and the live one on Kunavo, with native audio and transparent per-video pricing (Lite from $0.16, Fast from $0.32, Quality from $1.92 per 8s 720p clip). For the full model landscape — Veo, Sora, Kling and Hailuo — see the text-to-video API guide, and for image-to-video and the async task lifecycle the video docs.

FAQ