MiniMax Hailuo 3 (H3) API: How It Compares to Kling 3.0 Omni, Seedance 2.0, and Hailuo 2.3 on hiapi

MiniMax Hailuo 3 launched on July 31, 2026 — it's already live on hiapi

MiniMax teased the model under the #MiniMaxH3 hashtag on July 30, then shipped it the next day: MiniMax H3, the model the community still calls "Hailuo 3," succeeding Hailuo 2.3 as MiniMax's flagship video generator. It's an omni-modal transformer that takes text, images, video, and up to twelve reference files as input and returns native 2K video with synchronized stereo audio — dialogue, sound effects, and ambience generated in the same pass as the picture, no separate scoring or foley step.

The launch moved MiniMax's Hong Kong-listed stock (0100.HK) about 13% in a single session. It's a real, well-covered release, not a rumor — which is exactly why so many "Hailuo 3 API" searches are landing on comparison pages right now.

Here's the part that matters if you're trying to actually build with it: hiapi already has Hailuo 3 live, shipped as minimax-h3, at $0.118872 per second — noticeably below MiniMax's own $0.13/s Open Platform rate. Open weights haven't dropped yet (MiniMax said "within days" at launch; as of this writing there's still no MiniMaxAI repo on Hugging Face for it), so the API is the only way to use H3 today regardless of which provider you pick.

So the real question isn't "where do I find a Hailuo 3 alternative" — it's "given that hiapi has Hailuo 3, which of hiapi's four flagship video models should I actually route this job to." That's what this comparison answers, model by model, with the exact input schema and current per-second pricing for each.

The four contenders

Model	hiapi ID	Best for	Native audio	Max resolution	Price
MiniMax H3	minimax-h3	omni-modal scenes, dialogue-driven clips	Yes, native stereo	2K	$0.118872/s
Kling 3.0 Omni	kling-3.0-omni	resolution/audio control, tiered budget	Optional, toggled	4K	$0.10–$0.479/s (tiered)
Seedance 2.0	seedance-2.0	reference-video-guided edits, cheapest 4K	Optional (default on)	4K	$0.136–$1.486/s (tiered)
Hailuo 2.3	hailuo-2.3	fixed-price short clips, simplest schema	No	— (fixed-duration only)	$0.4 or $0.8 per video

Prices above are hiapi's current published rates as of August 2026 — pull /v1/pricing before you ship if you're budgeting a production run, since per-second video pricing on this platform has moved before (Seedance 2.0's 720p tier alone shifted from $0.33/s to $0.293/s between its own guide's worked example and today).

MiniMax H3 (minimax-h3) — the new flagship

This is Hailuo 3 itself. It's the model to reach for when the brief needs a complete scene — a character talking, ambient sound, music cue — generated in one call instead of stitched together from a silent clip plus a separate TTS pass.

{
  "model": "minimax-h3",
  "prompt": "A barista in a sunlit café explains today's roast to camera, warm ambient chatter and espresso machine hiss in the background",
  "duration": 15,
  "resolution": "2K"
}

At $0.118872/s, a 15-second 2K clip with native audio runs about $1.78 — for context, that's cheaper per second than Kling 3.0 Omni's 1080p-with-audio tier ($0.193/s) despite H3 running at a higher resolution.

Kling 3.0 Omni (kling-3.0-omni) — tiered control over resolution and audio

Kling 3.0 Omni is the pick when you need to dial resolution and audio independently to hit a budget, rather than accepting a model's fixed audio-always-on or audio-never behavior.

{
  "model": "kling-3.0-omni",
  "prompt": "Drone shot pushing through a neon night market, vendors calling out, crowd noise rising",
  "duration": 10,
  "resolution": "1080p",
  "aspect_ratio": "9:16",
  "sound": true
}

Sound is off by default — omit sound and you get a fully silent track, not a quiet one. The tiered pricing rewards turning it off when you don't need it: 720p silent is $0.10/s, 720p with audio jumps to $0.143/s; 1080p runs $0.129/s silent vs $0.193/s with audio; 4K tops out at $0.479/s (audio pricing not separately tiered at 4K in the current schedule).

Seedance 2.0 (seedance-2.0) — reference-video-guided edits, cheapest path to 4K

Seedance 2.0 is the only model of the four that accepts a reference video (reference_image_urls, up to 9 combined with first/last-frame inputs) rather than just a first-frame image — useful when you're extending or restyling existing footage rather than generating from scratch.

{
  "model": "seedance-2.0",
  "prompt": "Extend the pocket watch transformation sequence, gears reassembling in reverse, warm brass tones",
  "aspect_ratio": "16:9",
  "duration": 5,
  "resolution": "720p",
  "generate_audio": true
}

Audio defaults to on here (opposite of Kling Omni), and turning it off is the cheap lever: 720p with audio is $0.293/s vs $0.179/s with a reference video and audio off. The full tiered table: 480p $0.136/s ($0.082/s ref-guided), 720p $0.293/s ($0.179/s ref-guided), 1080p $0.729/s ($0.443/s ref-guided), 4K $1.486/s ($0.915/s ref-guided) — reference-guided generation is consistently cheaper across every tier, not just a discount at the low end.

Hailuo 2.3 (hailuo-2.3) — fixed price, three fields, no math required

Hailuo 2.3 (MiniMax's previous flagship, still live and still useful) is the simplest schema of the four and the only one priced per-video instead of per-second:

{
  "prompt": "Time-lapse of storm clouds rolling over a coastal cliff",
  "duration": "6",
  "prompt_optimizer": true
}

Note duration is a string — "6" or "10", not a number; sending 6 as an integer fails with a 400. There's no resolution or audio parameter to configure at all. A 6-second clip is a flat $0.40; 10 seconds is $0.80. If you're generating short clips at high volume and want predictable per-clip cost without tracking resolution tiers, this is the model that removes that variable entirely.

Which one to actually call

• Need dialogue or a fully-scored scene in one shot? minimax-h3. It's the newest model, the cheapest per-second 2K option here, and audio is native rather than bolted on.
• Need to hit a specific budget and don't mind toggling settings? kling-3.0-omni. The resolution/audio tiering gives you the most granular cost control of the four.
• Editing or extending existing footage? seedance-2.0. It's the only one of the four built around reference-video input, and the ref-guided discount makes it the cheapest way to reach 1080p or 4K if you already have source footage.
• High-volume short clips, fixed budget per clip, minimal config? hailuo-2.3. Three fields, flat per-video pricing, nothing to tune wrong.

All four run through the same async task API — submit with POST /v1/tasks, poll GET /v1/tasks/{id} until it reports a terminal status, then pull the output from output[0].url before the signed link expires. If you've already integrated one hiapi video model, switching between these four is a matter of swapping the model ID and adjusting the handful of fields above — see the Seedance 2.0 API guide or the Hailuo 2.3 text-to-video guide for full worked examples, or check current rates for all of them on the pricing page before committing to a production run.

Grab an API key from the dashboard and start with whichever model matches the job above — the schema differences are small enough that testing two of them side by side on the same prompt takes only a few minutes.