MiniMax H3 ref2v: Batch Social Video Content Variants

Key Takeaways

• MiniMax H3's ref2v takes images, videos, and audios arrays in one call, so several social-platform cuts can come from a single reference kit.
• The first 5 reference images per call are free; each one after adds $0.40 regardless of resolution.
• H3 has no seed field, so a winning take can't be replayed — variants come from changing the reference combination, not a locked seed.
• MiniMax H3 Max — a faster, fal Research–built variant — is announced but not yet on Siray; details below.

Introduction

This is not the launch-hero playbook — MiniMax H3 for E-commerce: A Product Launch Video Playbook covers one polished narrative cut around a single product story. This piece covers the opposite job: producing several social-content variants, different platform crops, pacing, and tone, from one shoot, by recombining the same reference kit. A product shot, a creator clip, and a background track go into ref2v's images, videos, and audios arrays; swapping which assets go into each call, not rewriting the prompt, turns one shoot into several usable cuts.

Coming Soon to Siray: MiniMax H3 Max

MiniMax H3 Max is a separate model from the MiniMax H3 above — a post-trained, speed-optimized variant built by fal Research on MiniMax's open-weights H3, not a MiniMax first-party release. It isn't on Siray yet.

Per fal's announcement, H3 Max renders a 5-second clip in under 3 seconds — about 35x base H3's throughput — and ranks #1 in fal's human-preference evaluations for quality, prompt understanding, and aesthetics against 12 models, including base H3, Wan 3.0, Seedance 2.5, Kling 3, and Veo 3.1.

The catch: a 768p resolution cap at launch, versus base H3's 2K/4K. Launch covers text-to-video and image-to-video only; ref2v arrived later. fal prices it at $0.08/second at 768p, 50% off the first 14 days.

Siray plans to add MiniMax H3 Max via the same unified API and one-key access as its other models; no launch date is set.

Why ref2v Fits Variant Batching

minimax/minimax-h3-ref2v accepts reference images, video, and audio in one call, with native stereo baked in — no separate audio pass per variant. The real lever isn't prompt text but which assets go in: drop the creator clip and keep the product shot plus audio, and output tilts product-forward; swap the background track and tone shifts without touching the prompt.

Building Variants from One Kit

ref2v requires aspect_ratio — 21:9, 16:9, 4:3, 1:1, 3:4, or 9:16 covers everything from a widescreen cutdown to a vertical Shorts crop — and duration, an enum of 5-15 whole seconds only, fitting a hook or short scene rather than a multi-beat story.

From there, vary the reference set per call: a vertical Shorts cut might lean on the creator clip and audio and drop the product still; a square feed cut might do the reverse. Each combination is a separate ref2v call against the same kit, not a new production.

What Extra Reference Images Actually Cost

Reference images aren't free past five: ref2v bills the first 5 per call at no extra charge, then $0.40 each, regardless of size. Base pricing is linear with no volume tiers — 768p $0.08/s, 2k $0.13/s, 4k $0.16/s.

Eight reference images in one 5-second 768p call runs about $1.60 — four times the cost at five images or fewer, and more than a same-length single-reference 4k call. Batching stays affordable when each call reuses a lean subset of the kit, not every asset.

No Seed Means No Replay

None of the three H3 endpoints expose seed, prompt_expansion_enable, or negative_prompt — a real limit for variants work. Treat a good result as the next call's reference source, not a seed to reuse: feed the clip back in, adjust the combination, and build forward from what's proven.

Developer Notes

MiniMax H3 runs on the same Siray API key as every other model. Teams already using other models — see One API for Multiple AI Models — can point existing code at ref2v directly. Full spec: MiniMax H3 (Hailuo 3.0), Explained.

Summary

Batching social video variants with MiniMax H3 comes down to one reference kit, recombined per call through ref2v, not a rewritten prompt per platform cut. The cost model rewards lean reference sets; the missing seed field means variants build forward from a good result instead of replaying one.

Get Started

Create your free Siray account and start batching social video variants with MiniMax H3's ref2v endpoint.