Tag

#video-generation

36 posts0 followers

Explore Hashnode

Alternatives

Callirra1d ago · 5 min read

Conditioning a video model with audio: three reference clips, and where lip sync actually stops

Quick answer: H3 accepts up to three reference audio clips (MP3 or WAV, 15 MB each) on image-to-video lanes from 480P to 2K, and never on 2K text-to-video. Audio conditions the result, but dedicated l

1 callirra.hashnode.dev

Callirra1d ago · 5 min read

Six style presets, one prompt: what drama staging and action staging actually change in MiniMax H3

Quick answer: minimax_h3_style_preset takes six values across two axes, staging (drama vs action) and treatment (live action vs animation, with an optional photoreal pass). Presets exist only on the m

1 callirra.hashnode.dev

Callirra1d ago · 5 min read

960P is the H3 tier nobody documents — and it runs 50% longer than standard 1080P for the same rate

Quick answer: 960P outputs 1728x960 at 3 credits per second, the same rate as 1080P, with a fifteen-second ceiling instead of standard 1080P's ten. Multimode 1080P also runs up to fifteen seconds. 960

1 callirra.hashnode.dev

Callirra1d ago · 5 min read

Two H3 tiers output 2560×1440. One costs 4 credits a second, the other 10

Quick answer: both lanes return 2560x1440, but 1440P bills 4 credits per second and 2K bills 10. The premium does not buy pixels: it buys video-to-video, six aspect ratios and first-plus-last-frame at

1 callirra.hashnode.dev

Callirra4d ago · 7 min read

What MiniMax H3 actually outputs: "1080P" is 1920×1056, durations snap to 24 fps

Quick answer: On the hosted MiniMax H3 lanes I tested, a "1080P" request comes back at 1920×1056, not 1920×1080 — the model converges to a pixel-area cap. Durations are quantised to whole 24 fps frame

3 callirra.hashnode.dev

Callirra4d ago · 6 min read

An AI video prompt pre-flight checklist: 12 checks before you spend a generation

Quick answer: Most wasted generations come from four decisions made before the model reads the prompt: an unclear subject/action split, reference media that fights the native aspect ratio, a duration

4 callirra.hashnode.dev

Henry Jin5d ago · 4 min read

What Telegram's compression really does to AI video (and the settings I use)

A few weeks ago someone on my Telegram channel said one of my AI clips looked "mushy." I'd exported it at 1080p, it looked sharp in the editor, and Telegram had quietly ruined it. Since I make most of

videoall.hashnode.dev

Kling4 AIOct 26 · 7 min read

A dependency-free pre-flight check for Seedance 2.5 reference images (TypeScript)

Seedance 2.5 can take up to 30 reference images in a single request, which is a big jump from Seedance 2.0's nine. More references also means more chances that one file breaks the request: a logo expo

ai-image-lab.hashnode.dev

Scofield EricOct 17 · 6 min read

How One-Photo AI Video Recasting Works (Using the Bulin 47 Trend as an Example)

This month's viral video format is a good case study in where consumer AI video is right now. The Bulin 47 trend takes one real street clip, a Dominican dembow artist freestyling outside a restaurant

scofieldfee.hashnode.dev

人民DOct 14 · 13 min read

AI Video Prompt Engineering: How to Turn a Single Image into a Cinematic Shot

Learn how to control camera movement, preserve character and product consistency, and avoid common image-to-video generation problems with practical prompting techniques. Creating an AI video from a s

wan3video.hashnode.dev

Search Hashnode

Explore Hashnode

Trending tags this week

Conditioning a video model with audio: three reference clips, and where lip sync actually stops

Six style presets, one prompt: what drama staging and action staging actually change in MiniMax H3

960P is the H3 tier nobody documents — and it runs 50% longer than standard 1080P for the same rate

Two H3 tiers output 2560×1440. One costs 4 credits a second, the other 10

What MiniMax H3 actually outputs: "1080P" is 1920×1056, durations snap to 24 fps

An AI video prompt pre-flight checklist: 12 checks before you spend a generation

What Telegram's compression really does to AI video (and the settings I use)

A dependency-free pre-flight check for Seedance 2.5 reference images (TypeScript)

How One-Photo AI Video Recasting Works (Using the Bulin 47 Trend as an Example)

AI Video Prompt Engineering: How to Turn a Single Image into a Cinematic Shot