Post

https://twitter.com/1582819328281481225/status/1850251733399703826

This Thread may be Removed Anytime!

Twitter may remove this content at anytime! Save it as PDF for later use!

More from @AngryTomtweets

Angry Tom

@AngryTomtweets

Dec 18

Oh my... Kling just dropped the next era of motion control.

Kling VIDEO 2.6 can copy any action with perfect lip-sync, lifelike motion and expressive gesture.

It outperforms Wan 2.2-Animate, Act-Two and DreamActor 1.5 across all metrics.

More examples below.

Based on internal tests, Kling VIDEO 2.6 Motion Control outperforms all leading models, including Wan 2.2-Animate, Runway Act-Two, and DreamActor 1.5.

1. Perfectly Synchronized Full-Body Motions

Capture body movements, expressions, and lip sync in full detail, seamlessly synchronized with background sounds for a smooth final output.

Read 8 tweets

Angry Tom

@AngryTomtweets

Dec 3

RIP "silent" models.

Kling just dropped their first native audio model, VIDEO 2.6 and it's insane.

Previously, Kling's video models could only generate "silent visuals".

But now... native audio, insane realism, stronger understanding and full audio control.

More examples below.

This update introduces a groundbreaking "Native Audio" capability. The model completely transforms the traditional AI video workflow of "first generating silent visuals, then manually adding voiceovers and sound effects."

By deeply aligning the semantics of sounds and dynamic visuals from the physical world, VIDEO 2.6 enables the end-to-end generation of complete videos in a single go...

1. Text-to-Audio-Visual

Simply input text to generate a video that includes voice, sound effects, and ambient sounds.

Prompt: A young asian woman, casually dressed, sitting on a sofa in a cozy living room, softly saying: “I have a secret, Kling 2.6 is coming.”

Read 10 tweets

Angry Tom

@AngryTomtweets

Sep 30

This is crazy...

Open AI just dropped Sora 2, the best AI video model in the world.

It handles physics and motion like no other model while also enabling realistic, controllable video with sound.

Here are 10 insane examples:

Sora 2 can do things that are exceptionally difficult and in some instances outright impossible for prior video generation models...

1. Olympic gymnastics routines....

Read 13 tweets

Angry Tom

@AngryTomtweets

Sep 18

AI is getting crazier...

Higgsfield just dropped Lipsync Studio, an all-in-one space for creating the most realistic lip-sync videos yet.

Unlimited Kling Speak in 720p, Kling Lipsync and InfiniteTalk until September 22.

6 insane examples:

Read 7 tweets

Angry Tom

@AngryTomtweets

Sep 9

There is no way to tell anymore.

Nano Banana is NOW LIVE alongside FLUX Premium in LTX Studio.

Two of the best image models on the market right now.

Here's what's possible now: (with examples)

1. What’s the difference?

- FLUX Premium is optimized for high-fidelity image generation.
- Nano Banana is a state-of-the-art image editing model.

Here's how it works...

Start with a cinematic portrait generated using FLUX Premium and you'll get sharp facial detail, natural lighting, and stunning realism.

Read 7 tweets

Angry Tom

@AngryTomtweets

Aug 30

omg... this is crazy!

Higgsfield just unlocked Start & End Frame in 1080p powered by MiniMax.

Set the first shot, set the last, it builds everything in between.

Cinematic camera moves and VFX stacked with preset.

5 examples: