Native 2K with single-pass audio
Dialogue, sound effects, and ambient audio all generate in the same pass as the video — no separate audio stage, no post-hoc voiceover step. That's a direct upgrade over Hailuo 2.3, which has no native audio at all.
Hailuo 3.0 — officially named MiniMax H3 — is MiniMax's next-generation AI video model, released July 31, 2026, that generates native 2K video with synchronized dialogue, sound effects, and ambience in a single pass, up to 15 seconds per clip, with an extensive Omni-Reference system for character and style consistency.
Built for video creators, marketers, and studios, it's a direct upgrade from Hailuo 2.3 — reading text, images, video, and audio as one unified context, so reference-driven generation and instruction-based editing work as core features rather than add-ons.
Dialogue-driven scenes, product ads, short films, music videos, character-consistent series
Video and marketing teams that need native audio, precise reference control, and editing without regenerating
Native 2K, native audio, and deep reference control — in a single generation pass.
Dialogue, sound effects, and ambient audio all generate in the same pass as the video — no separate audio stage, no post-hoc voiceover step. That's a direct upgrade over Hailuo 2.3, which has no native audio at all.
Supply up to 9 images, 3 video clips, and 3 audio clips in one generation to guide character appearance, visual style, motion, and voice — the reference depth that makes recurring characters and matched brand style practical.
Revise a character, object, scene, or line of dialogue on an existing clip using plain instructions, without regenerating the whole thing from scratch — a targeted fix instead of a full re-roll.
Multiple shots can sit inside a single generation — a title sequence or a shot-reverse-shot dialogue beat produced in one run rather than cut together from separate clips.
MiniMax released Hailuo 3.0's base weights on August 3, 2026, just days after launch — one of the first frontier-grade video models to ship with an open-weight commitment, under a planned license covering commercial use for smaller organizations.
Hailuo 3.0 fits wherever teams need native audio, precise reference control, and editing without regenerating from scratch.
Add a node, describe your scene, and generate native 2K video with audio in one pass.
Add a Hailuo 3.0 node to the canvas and describe your idea in a text prompt, or upload a first-frame image. Optionally add Omni-Reference inputs — up to 9 images, 3 video clips, and 3 audio clips — for character, style, or voice consistency.
Set duration (5–15 seconds) and aspect ratio. Native audio — dialogue, SFX, and ambience — generates in the same pass as the video. Multiple shots can sit inside a single generation.
Use instruction-based editing to revise a character, object, or line of dialogue on an existing clip without regenerating from scratch, chain into an upscaler, then export with commercial rights on your plan.
Real generated videos with the exact prompts — copy any prompt and run it yourself.
One subscription across 30+ AI models — no per-tool credit balances or separate signups. Credit cost per generation is shown live in the node before you run it.
500 weekly credits to test the model and see the output quality. No credit card required.
45,000 monthly credits with cheaper per-credit pricing, commercial rights, and video download.
90,000 monthly credits, up to 10 seats, shared workspace, centralized billing, and priority support.
210,000 monthly credits, API and integration support, unlimited seats, and dedicated management.
Full specs for Hailuo 3.0 (MiniMax H3).
How Hailuo 3.0 compares with other AI video models in the Phygital+ catalog.
Chain image, video, upscaling, and audio into one repeatable workflow.
Design a reference frame with an image model, then animate it with Hailuo 3.0 in one workflow.
Generate with Hailuo 3.0, then upscale to delivery resolution for social, ads, or broadcast.
Layer extra music or sound design on top of Hailuo 3.0's native audio with a dedicated node.
Write the script and shot list with a text model, then feed it straight into video generation.
Everything you need to know about Hailuo 3.0 in Phygital+.
Hailuo 3.0 — officially named MiniMax H3 — is MiniMax's next-generation AI video model, unveiled July 17, 2026 at WAIC and released July 31, 2026. It's an omni-modal model that reads text, images, video, and audio as one unified context, generating native 2K video with synchronized dialogue, sound effects, and ambience in a single pass, up to 15 seconds per clip.
Usage is credit-based and included in every Phygital+ plan — from the free plan up to Enterprise. The exact credit cost per generation is shown live in the node before you run it. This is separate from MiniMax's own pay-as-you-go API pricing, reported at roughly $0.13 per second of 2K video — note that as a very new model, MiniMax's own pricing may still change.
Outputs created on any paid plan (Starter and up) come with commercial-use rights. Separately, MiniMax has released Hailuo 3.0's weights under a planned MiniMax Community License permitting commercial use for organizations under $20M in revenue, with attribution — check current license terms if you plan to self-host. Check the model card for any provider-specific limits.
Standard generations run 5 to 15 seconds, with multiple shots able to sit inside a single clip — a title sequence or a shot-reverse-shot dialogue beat can be produced in one run instead of stitched from separate generations. Resolution is native 2K (1440p on the short edge) at a fixed 24fps, across six aspect ratios including 16:9, 9:16, and 21:9.
Omni-Reference is Hailuo 3.0's reference system: supply up to 9 images, 3 video clips, and 3 audio clips in a single generation to guide character appearance, visual style, motion, and voice consistency. It's what makes recurring characters and matched brand style practical across a series of generations.
Yes — this is one of Hailuo 3.0's headline features. Dialogue, sound effects, and ambient audio all generate in the same pass as the video, with no separate audio stage. This is a genuine upgrade over Hailuo 2.3, which does not generate native audio at all.
Yes — MiniMax released Hailuo 3.0's base weights on August 3, 2026, days after the model's public launch, making it one of the first frontier-grade video models with an open-weight release. Licensing is planned under a MiniMax Community License covering commercial use for smaller organizations; verify current terms before self-hosting, since this is a very recent and still-evolving release.
Standard Hailuo 2.3 tops out at 1080p for 6 seconds (or 768p for 10 seconds) with no native audio and no reference system. Hailuo 3.0 is a substantial generational jump: native 2K, 15-second single-pass clips with built-in dialogue and sound, Omni-Reference for character and style consistency, and instruction-based editing on existing clips.
Yes. MiniMax offers Hailuo 3.0 through its Open Platform API on a pay-as-you-go basis, alongside the Hailuo app, the MiniMax Hub desktop app, and third-party hosts. In Phygital+, you can run it as a node on any plan, with API and integration support on the Enterprise plan.
Running Hailuo 3.0 in Phygital+ means no separate signup or credit balance for one model: it sits on the same canvas as 30+ other AI tools. Chain generation, instruction-based editing, upscaling, and other models into one repeatable workflow under a single subscription.
Browse the full catalog of 30+ AI models available in Phygital+.
Join 100+ teams using Phygital+ – every model in one workspace
Try Phygital+ free