Hey dude, Text2video works for me. Uploading faces don't. Sometimes character sheets work too. The sheet can't have a close up though. Or just draw an X or some lines over the faces and try it that way. It sometimes breaks the filter. I ranted about dreamina last week and some said it depends on the region. I can't extend videos anymore on dreamina which have faces. They got stricter I think. It is all so stupid man. Such nonsense.

Well, at the time of the post they did not released v2v endpoint in Fal. I saw only text2video and image2video. Now, I'm seeing video edit and reframe but still multiple keyframe parameter is missing 👀 @ailker @jfischoff

yoooo text2video

"Prototype" Music: Stable Audio 3 text2music Video: LTX 2.3 audioreactive text2video 100% created with local open source models #StableAudio3 #LTX2 #ComfyUI

How are you getting any of these character images through the stringent Higgsfield IP filters. Are you using character sheets or pure text2video?

Missing the unique charm of OSS Text2Video vibes from pre-Sora days? We're open-sourcing StreamingT2V, an autoregressive method that creates dynamic & consistent long videos -> now 1200 frames or 2 full minutes :) code/paper: github.com/Picsart-AI-Resear… Kudos to @PicsartAI team!🎉

Less than 3 years ago, text2video barely existed. Today we have Sora, Veo, Kling, Runway, Seedance and half a dozen other serious contenders fighting for the crown. Every few months, video quality takes another leap. Characters stay consistent longer. Physics improve. Camera movements become cinematic. Editing gets cleaner. The gap between AI video and traditional production gets smaller every quarter. Remember when a 5-second AI clip looked impressive? Now people scroll past 30-second AI scenes without realizing they're AI. That's how quickly the baseline moved. And while the public watches model releases, something bigger is happening behind the scenes. Scorsese working with Black Forest Labs isn't random. It signals that Hollywood talent is already entering the ecosystem. You think Google, OpenAI, and the rest aren't building similar relationships? They are. The announcements just haven't dropped yet.

なんてこった！ちょっとこの二週間くらい休みなく働いていて疲れが出てきたのか、Seedance2.0の無駄遣いをしてしまった。 多分Seedance2.0はちゃんとリファレンスやファーストの画像は用意したほうが良さそう。 これはText2Videoですぜ😎

do it. Grok Imagine Text2Video #nsfw #aigirls #aiart #aigrok #GrokImagine #GrokImaginespicy #grokporn #grokmakeporn #grokvideo #aivideo #aiporn #GrokImagineNSFW #GrokNSFW

I'd definitely recommend these ones as must haves: IC-LoRA Outpaint by oumoumad - (turns 4:3 img/vid into 16:9) Edit Anything by Alissonerdx - Still a WiP, but can edit most things in a video using a prompt SYSTMS-FLW-IC-LORA by systms (see video)- Allows for much more aggressive/dynamic/fluid scene transitions between two or more videos (can be used for looping too) and even works with img2vid by inserting a gray colored picture between two different images using a first-middle-last img2vid guide node (ComfyUI). This LoRA absolutely blows away the default LTX first-middle-last frame capabilities, which only work well when the two images are pretty closely resembling each other. If they are too different, LTX just hard fades or cuts directly to the second image scene, which is super boring. I want the scene to actively morph and change from image to image (even if it's unrealistic). For the video I used 5 images, 3 reference images with 2 gray images inserted between them (241 frames total, so each one was offset by 60 frames: 0, 60, 120, 180 and -1 for the last image frame). And it blends them together much better than the default way LTX does it. It's not perfect, but it's definitely fun to play around with. I miss being able to have this kind of aggressive multi-image transition that I used to get with the WAN 2.1 5b fun inpaint model, and this LoRA brings some of that functionality to LTX 2.3. I also didn't direct the video at all with the prompt, I just described the character and added no instructions for motion, movement etc. It did it all on it's own. OmniNFT-RL-Lora by Kijai - This is just a nice LoRA that adds scene details and makes your text2video videos look so much better (and helps with audio too). Doesn't seem to do anything with I2V, but is a great LoRA to just turn on by default when doing T2V, as it will basically improve everything with no downsides (that I know of). I stack it with the 19b IC detailer LoRA to really sharpen thing up. LTX-2 VBVR LoRA by LiconStudio - This one basically aims to improve LTX 2.3's ability to do physics and reasoning more accurately (falling objects, object trajectory, object collision etc). Works only with img2vid. All of these are on HuggingFace.

nice view... Grok Imagine Text2Video #nsfw #aigirls #aiart #aigrok #GrokImagine #GrokImaginespicy #grokporn #grokmakeporn #grokvideo #aivideo #aiporn #GrokImagineNSFW #GrokNSFW

Thinking about it... Grok Imagine Text2Video #nsfw #aigirls #aiart #aigrok #GrokImagine #GrokImaginespicy #grokporn #grokmakeporn #grokvideo #aivideo #aiporn #GrokImagineNSFW #GrokNSFW

To the victor go the spoils. Grok Imagine Text2Video #nsfw #aigirls #aiart #aigrok #GrokImagine #GrokImaginespicy #grokporn #grokmakeporn #grokvideo #aivideo #aiporn #GrokImagineNSFW #GrokNSFW

Space Marines 1 Lost some of my armor in the battle, Sir...can you check for damage? Grok Imagine Text2Video #nsfw #aigirls #aiart #aigrok #GrokImagine #GrokImaginespicy #grokporn #grokmakeporn #grokvideo #aivideo #aiporn #GrokImagineNSFW #GrokNSFW

Space Marines 2 This new armor is a lot more comfortable, Sir... Grok Imagine Text2Video #nsfw #aigirls #aiart #aigrok #GrokImagine #GrokImaginespicy #grokporn #grokmakeporn #grokvideo #aivideo #aiporn #GrokImagineNSFW #GrokNSFW

CANVAS REIMAGINED INTERNATIONAL QP Theme:Unbound Glow Host: the amazing @NecoHashimoto Text2Video made with @SocialSight Come join us: @JohnWalkerCXD @JuC69 @Stormbourne_76 @l_beautiful1

test text2video Grok Imagine

I have an AI CHALLENGE for you! 🔥 ONLY BY PROMPTING Text2Image or Text2Video create THIS effect: A person pressing his/her face agaist glas. Bonus points if it is against the screen of the video itself. Show your Skill 🏆OR your Fails 😅 Please share the result and Prompt below. You will win.... ... nothing. This is for the love of the game. 🥰

In space no one can hear you cream. Grok Imagine. Text2Video #GrokImagine #NSFW #nsfwai #grok #aivideo #grokimaginensfw