Model updates, pricing, and access
Veo 3.1 Lite/Fast releases, capabilities, API and AI Studio availability, pricing, clip formats, and global product rollouts.
45.2%
Best tweets about Google Veo
42 Best Tweets About Google Veo (2026)
Find the best tweets about Google Veo, covering video model releases, prompts, image-to-video, filmmaking, benchmarks, access, and creator experiments.
Google Veo video generation, model updates, prompting, visual quality, access, comparisons, filmmaking workflows, and demonstrated outputs.
- Creators
- 37
- Updated
Original Xholic analysis
What 42 top Google Veo posts reveal
Discussion centers on lower-cost Veo 3.1 Lite access and controllable production workflows. Tutorials had the highest median all-time score among formats, while several Veo 3.1 Lite announcements were score outliers. Posts also include both visually compelling demonstrations and cautions about output consistency and scientific fidelity.
- Dominant tone
- Positive
- Median score
- 14.4
- Leading format
- Announcement
- Recent posts
- 9.5%
83.3% of posts
All-time engagement
50% of posts
Published in 90 days
Conversation map
The themes creators return to
Filmmaking and demonstrated creative outputs
Veo-made short films, vlogs, historical FPV sequences, product cinematography, logo motion, memes, and experimental visual storytelling.
26.2%
Prompting and generation controls
Detailed cinematic prompting, camera direction, image references, first/last frames, multi-image inputs, audio, aspect ratios, and output consistency techniques.
26.2%
Visual realism, identity, and consistency
Photorealism, avatar direction, self-cloning, character consistency, lip sync, natural motion, and limitations or artifacts in generated video.
26.2%
Google ecosystem integrations
Veo embedded in Flow, Google Vids, Google Ads, YouTube Effect Maker, Gemini, and developer tooling for broader creative production.
23.8%
Creator, ad, and influencer workflows
Repeatable pipelines using Veo with image models, scripts, voice, editing, and product assets for AI influencers, UGC-style ads, podcasts, and scalable marketing creative.
16.7%
Developer builds and API workflows
Gemini API access, OpenAI-compatible endpoints, and apps built around Veo generation, interpolation, extension, storage, and video post-processing.
16.7%
Quality comparisons and benchmarks
Comparisons with Kling, Seedance, SkyReels, Sora, and other video models across realism, speed, cost, rankings, and scientific correctness.
14.3%
Tone and stance
Sentiment Positive leads
Author posture Supportive leads
Performance benchmark
- Median likes
- 43
- Median reposts
- 4
- Median replies
- 4
- Median views
- 4.7K
Posts with media make up 73.8% of this collection. Their median all-time score is 10.2, compared with 65.1 for text-only posts.
Format mix
- Announcement 50% · score 8.29
- Case Study 26.2% · score 18.7
- Opinion 11.9% · score 8.20
- Tutorial 9.5% · score 127.3
Consensus and debate
Where creators agree, and where they do not
Shared view
Veo 3.1 Lite supports iterative workflows
Posts describe Veo 3.1 Lite as a lower-cost option for text-to-video and image-to-video generation. Flow posts specifically position it for fast prompt iteration and storyboarding before final output work.
Shared view
Control comes from structured inputs
Practical guidance emphasizes reference images, detailed character and camera descriptions, first/last frames, and restrained motion prompts to improve continuity and directability.
Shared view
Veo appears within broader creative workflows
Posts describe Veo alongside Flow, YouTube Effect Maker, image models, API access, external voice and editing tools, and custom applications rather than solely as a standalone generator.
Shared view
Demonstrations span cinematic and commercial concepts
Shared examples include product cinematography, historical FPV sequences, daily-vlog concepts, memes, logo animation, and hybrid design work, showing a broad range of creative-use framing.
Open debate
Visual persuasion versus technical fidelity
Some posts describe outputs as visually compelling or hard to immediately identify as AI, while others identify visible generation inconsistencies or limitations in scientific understanding despite photorealistic appearance.
Open debate
Veo’s competitive standing varies by benchmark and task
One post reports Veo 3.1 Fast and Veo 3.1 in first and second place for multi-image-to-video, while another reports SkyReels V4 ahead of Veo 3.1 in a text-to-video-with-audio arena. Another post expresses demand for a Veo 4 update.
What performs
Patterns behind standout posts
Tutorials were the engagement standout
Tutorials had a median all-time score of 127.29, above announcements at 8.294 and case studies at 18.67. The leading outlier was a self-cloning workflow post with an all-time score of 1109.89.
Several launch posts were score outliers
The supplied outliers include the Veo 3.1 Lite cost-efficient-model announcement at 350.71, API/AI Studio availability at 155.95, and the wider Labs update at 389.05.
Media was common, but had a lower supplied median score
Media appeared in 31 posts, or 73.8% of the set. The supplied median all-time score was 10.197 for media posts, compared with 65.073 for text posts.
Statistical standouts
- View standout post 1 Score 1109.9 · 76.92× median
- View standout post 2 Score 389.1 · 26.96× median
- View standout post 3 Score 350.7 · 24.3× median
- View standout post 4 Score 155.9 · 10.81× median
- View standout post 5 Score 138.3 · 9.59× median
Creator landscape
Who shapes this conversation
The five most represented creators account for 23.8% of the selected posts.
-
1. Philipp Schmid
@_philschmid
2 posts
-
2. 👩💻 Paige Bailey
@DynamicWebPaige
2 posts
-
3. Julian Goldie SEO
@JulianGoldieSEO
2 posts
-
4. Shushant Lakhyani
@shushant_l
2 posts
-
5. DEV Community
@ThePracticalDev
2 posts
-
6. 0x ROAS
@0xROAS
1 post
Workflow posts offer repeatable production approaches
Creator posts outline reusable pipelines for selfie-character generation, image-to-video influencer production, and sequential ad assembly using frame references and editing.
Developer posts focus on access and implementation
Developer-oriented posts highlight Gemini API and AI Studio availability, OpenAI-compatible endpoints, and an application that generates, interpolates, extends, and stores Veo videos.
Ecosystem posts position Veo across creator and advertiser products
Posts place Veo in Flow, Google Vids, YouTube Effect Maker, and Google Ads, including creator tools and advertiser-facing video generation.
Since the previous snapshot
What changed since Aug 12, 2026
- 78.6% of the selected posts remained.
- The creator count changed by +1.
- The leading sentiment remained stable.
How this analysis was made
Themes, sentiment, stance, and post format are classified per tweet. All counts, shares, medians, creator concentration, freshness, and performance comparisons are then calculated directly from the published snapshot.
Xholic's all-time score compares engagement while accounting for reach, post age, and creator consistency. It is used for relative comparisons within this collection.
This report analyzes the exact 42-post snapshot shown below. AI identifies editorial categories and drafts explanations; all statistics are calculated from the snapshot, and every narrative claim is checked against cited posts before publication.
Best Google Veo tweets
Top Google Veo tweets from 37 creators
Ranked 01–42
- 01
@EXM7777 ·
Google Veo 3.1 lets you clone yourself from a single selfie and produce daily content without ever recording a video... here's the workflow: STEP 1: analyze your selfie with Gemini - upload a photo of yourself - prompt: "analyze this photo and write a detailed Veo 3 character description including facial features, hair, skin tone, clothing, and body language" - Gemini outputs a character profile STEP 2: generate video prompts - take that character description - ask Gemini: "write a 100-150 word cinematic Veo 3 prompt with character appearance, camera angle, and dialogue in quotes" some key tricks: start with "a selfie video of...", make the arm visible (sells the selfie illusion), add natural eye movement, put dialogue in quotes (Veo generates synced audio) and specify "no subtitles" STEP 3: generate with Veo 3.1 - paste into Google AI Studio - use "Ingredients to Video" : upload 4 reference photos for face consistency - output: 8-second clips at 1080p with synchronized dialogue you can create a lot of formats with this... the one going most viral: selfie-style walking + talking to camera... viewers genuinely can't tell it's AI
- 02
@GoogleLabs ·
Woah. It’s been a little over 2 months since the start of 2026. Here's what we’ve dropped so far: 🚀 @FlowbyGoogle: Our biggest update yet. We’ve redesigned the interface, bringing image generation to the forefront, making it easier to manage assets, and giving you more precise control over your images and videos. We also made improvements to Veo 3.1 and welcomed Nano Banana 2 to the model lineup for even more expressive and consistent outputs. (Also, come say hi on Instagram: @ flowbygoogle!) 🚀 @JulesAgent: A major speed boost, all Jules free-tier users have been upgraded from Gemini 2.5 Pro to Gemini 3 Flash. 🚀 SynthID: We've added the ability to check whether audio is Google AI-generated (in addition to image and video). Generative media content made with Labs products can be uploaded to the Gemini app to see if they've been generated using Google AI. Simply type @ SynthID and ask "Is this AI?" 🚀 @StitchbyGoogle: We leveled up the MCP ecosystem. Users can now get step-by-step MCP client instructions and grab their API key directly from the Exports panel. We also joined the Antigravity MCP store and added new tools, like a coding agent that can ask to edit screens and generate screen variants. 🚀 Project Genie: Introducing an early research prototype powered by @GoogleDeepMind’s Genie 3 model, Project Genie lets users create and explore infinitely diverse worlds. Available to AI Ultra subscribers in the US. 🚀 @NotebookLM: A massive month for mobile. Users can now add customization prompts to Infographics and Slide Decks directly and generate video overviews on the mobile app. Plus, NotebookLM partnered with @Zillow to launch a featured notebook pre-loaded with their extensive home buying resources. Be on the lookout for other exciting partnerships to come! 🚀 Pomelli: Launched ‘Photoshoot,’ allowing users to take a single image of their product and easily create high quality, customized product shots to elevate their marketing. 🚀 Opal: Introduced a new agent step that analyzes a goal, determines the next best step, and automatically calls models and tools to finish a the task, such as Veo for video or web search for research. 🚀 @Producer_AI: Welcomed ProducerAI to Labs! ProducerAI is a creative collaborator, whether writing lyrics, refining a melody, or inventing entirely new genres. What a start to the year. 2026 is just getting warmed up. Explore all of our experiments at: https://t.co/TcBfiPohNK
- 03
- 04
@googleaidevs ·
Introducing Veo 3.1 Lite, now available via the Gemini API and @GoogleAIStudio. Veo 3.1 Lite supports both Text-to-Video and Image-to-Video, and is less than half the cost of Veo 3.1 Fast.
Watch video
- 05
@alexcooldev ·
The simplest way to create an AI influencer using workflow (3 steps): Nano Banana → Veo 3 → Arcads AI Workflow Step 1: Build your character with a JSON prompt (Nano Banana Workflow) → Use a structured JSON prompt instead of writing random text → Define the look clearly: character, outfit, lighting, camera style, background → Add negative constraints like: no logos, no text, no extra fingers, no CGI skin, no warped mirror This makes the output more stable and reusable. You’re not just prompting. You’re building a repeatable image system Step 2: Turn image → video (Veo 3 Workflow) → Connect the Nano Banana output directly into Veo 3 → Add a simple motion prompt: slight body movement, natural head turn, subtle phone movement → Keep it minimal so it feels more real Workflow removes the manual back-and-forth. Step 3: Add script + voice (Arcads AI Workflow) → Feed the generated video into Arcads AI → Add a short script, usually 10–20 seconds → Generate voice + lipsync that matches the character vibe Now the character becomes actual content. Why workflow matters: Prompt → Image → Video → Voice No messy manual process. No repeating the same setup every time. Just a clean pipeline you can scale. Nano Banana + Veo 3 + Arcads AI + workflow feature = simple AI influencer factory
Watch video
- 06
- 07
@0xROAS ·
this AI PODCAST VIDEO cost less than $1. VEO 3.1 is literally crazy cost efficient and one of my fav models rn. you can do anything you want: - talking videos - podcasts - 3d pixar animations - claymation - whatever you want and literally the cost per 8s video is $0.05 p.s i know voice can be fixed easy
Watch video
- 08
@cjfeth ·
Here’s how to make ultra realistic AI video ads that print: 1. Generate or find your image first. We search on Google/TikTok 2. Upload it to higgsfield (image) & we make customizations such as changing the outfit, background etc. 3. Once we have our final image, we move to Veo to generate the clips we need through different prompts (use Gemini to help you with these) Example: Generate a POV/selfie-style video of a female dermatologist using this image as reference. Maintain mid-chest-to-head framing with casual, handheld iPhone realism. She reaches forward from in front of the camera, naturally picks up the product, and holds it in one hand while explaining. As she speaks, she slowly walks forward, creating gentle forward motion with subtle camera shifts and natural handheld sway, including minor background parallax as if the phone is held while walking.' 4. Input your script for this scene 5. Keep repeating to generate all of your clips. To keep it consistent every time you generate a clip, grab the last frame and use it as a reference image for the next clip. 6. Generate all clips, in final editing you can use b-roll that help the transitions between AI generated clips look more natural. 7. Launch & print Authority figure AI ads are taking over 90% of my spend right now so don’t sleep on it
- 09
@nealmohan ·
Here’s what we launched at @YouTube this month: ⬆️ Increased thumbnail file size (up to 50MB from 2MB) 🌎 Expanded access to localized thumbnails 🛍️ Expanded access to our Shopping Affiliate program 🪄 Expanded access and Veo 3 integration in Effect Maker, a web-based tool that allows you to build effects for other @YouTubeCreators 🤗 A $20M global teen wellbeing initiative with @Googleorg 🤝 A unified advertiser and creator suite: YouTube Creator Partnerships 🔍 A new grid of Shorts on search results on mobile 💎 Gifts and Jewels on vertical live streams in 🇰🇷 🍌 @NanoBanana in our YouTube Create app 🛡️ Expanded likeness detection for government officials, journalists and political candidates ⚽️ A partnership with @FIFAcom for @FIFAWorldCup
- 10
@ArtificialAnlys ·
SkyReels V4 takes the #1 spot in Text to Video With Audio in the Artificial Analysis Video Arena, surpassing Kling 3.0 and Veo 3.1! SkyReels V4 is the latest video generation model from @Skywork_ai, marking a major shift from their previous avatar-focused models to a full multimodal video generation system supporting Text, Image, Video, and Audio inputs. The model generates up to 15-second videos at 1080p resolution with native audio support. SkyReels V4 also performs strongly across other modalities, ranking #2 in Text to Video without Audio, #4 in Image to Video with Audio, and #7 in Image to Video without Audio. The model is priced at $7.20 per minute with audio and $8.40 per minute without audio, positioning it below Kling 3.0 1080p Pro (~$20/min with audio) and Veo 3.1 ($24/min with audio), though at a premium over Grok Imagine at $4.20/min with audio. SkyReels V4 is available via the @SkyReels website, with both a web app and API access. SkReels V4 Omni will be released soon. See below for example generations of SkyReels V4 in the Artificial Analysis Video Arena 🧵
- 11
@FlowbyGoogle ·
Veo 3.1 Lite is now available in Flow. It’s perfect for fast prompt iterations and storyboarding before upscaling to the final product. Here's how you can use it: — Text and Image to Video — Landscape (16:9) and Portrait (9:16) — First and Last Frame We can't wait to see what you create!
- 12
@ahuja_priyank ·
🚨BREAKING: New AI capabilities are arriving in Google Vids, including high-quality video generation powered by Veo 3.1, available at no cost. Anyone with a Google account can bring stories to life from a simple prompt or photo. Check-out new features:
Watch video
- 13
- 14
@_philschmid ·
Veo 3.1 and Gemini image models are now available through the OpenAI compatibility layer, no SDK swap needed. 🎬 Generate videos via `/v1/videos` using Veo 3.1 🍌 Generate images with Nano Banana via `images.generate` 🔌 Drop-in compatible with OpenAI Python and JS SDKs 🔀 Switch to Gemini API by changing just 3 lines of code
- 15
@DesignArena ·
BREAKING: Veo 3.1 Fast and Veo 3.1 by @GoogleDeepMind are in 1st and 2nd place on Multi-Image to Video Arena These models can successfully reference multiple input images to create a video that users love At an average generation time of 48 seconds, they are also the two fastest video generation models Huge congrats to the @GoogleDeepMind team for this achievement!
- 16
@_philschmid ·
Veo 3.1 Lite now available in Gemini API and @GoogleAIStudio. Designed for rapid prototyping and high-volume video generation, starting at $0.05/sec. 🪶 - 1/2 the cost of Veo 3.1 Fast. - Text-to-Video (T2V) & Image-to-Video (I2V). - Landscape (16:9) and Portrait (9:16) format - 4s, 6s, and 8s clips.
- 17
@gethookdai ·
$4.3M in 90 days. Here's the AI ad system behind it. I stopped paying agencies $340 per creative. Now I spend $0.83 per finished ad and produce 10x more. THE STACK: - Gethookd for competitor research. Every Monday I pull every active ad from the top 50 spenders in my niche. 3 minutes. I'm looking at hook style, format, offer structure, and how long each ad has been running. Anything active 60+ days goes into my swipe file. - Claude for script generation. I feed it the top 5 hooks from my swipe file + my product USPs and ask for 20 variations. Prompt: "Write 20 direct response hooks for [product] using the pain-agitate-solve framework. Each hook must be under 12 words. Reference these winning hooks as inspiration: [paste hooks]." 20 scripts in 90 seconds. - Flux Kontext for product compositing. Upload your actual product photo and it places it into any scene while keeping the label pixel-perfect. No photographer. No studio. $0.15 per image. - Kling AI for video generation. Upload a reference image + audio script. Generates a talking head video with lip sync for $0.60 per clip. 5-second and 15-second variations of every hook. - Veo 3 for product-in-motion shots. Feed it a Flux composite and it animates the product. Someone pouring your supplement into a smoothie. $0.60 per clip. Previously cost $2,000 per product video shoot. EVERY MONDAY: 1. Pull top 50 competitors' active ads from Gethookd (3 min) 2. Identify top 10 hooks by longevity — anything running 60+ days (5 min) 3. Generate 20 hook scripts via Claude (2 min) 4. Generate 20 product composites via Flux Kontext (8 min, $3 total) 5. Generate 20 video clips via Kling + Veo 3 (15 min, $24 total) 6. Assemble with ffmpeg — hook + body + CTA + product shot (10 min) 7. Batch upload 20 finished creatives into Advantage+ (5 min) 48 minutes. $27. My old agency charged $6,800 for the same 20 creatives. TESTING PROTOCOL: - Each creative gets $25/day in its own ad set - Kill anything above $2.50 cost-per-ThruPlay after 48 hours - Kill anything below 1.8% outbound CTR after 72 hours - Kill anything above target CPA ($65) after $130 spend - Average hit rate: 3-4 winners out of 20. Those 3-4 run for 6-8 weeks and generate $400k-$600k in revenue each $27 in AI credits. $4.3M in 90 days. That's the math now.
Watch video
- 18
@zazzygfx ·
been playing with “daily vlog moments” concepts using Veo and Seedance within Flora focusing on small human moments instead of big ideas with a bit of voiceovers this came out really interesting 👇 prompt + workflow below
Watch video
- 19
@LinusEkenstam ·
I made the realistic trollface meme with Google Veo 3 and Nanobanan Pro in late November for fun. In the last week it picked up over 500.000.000 views combined on TikTok amongst the late Gen-z generation. “told teach to check email”
Watch video
- 20
@alex_barashkov ·
Probably our cringiest branding and design work yet. The console was rendered in Blender. The character’s face came from Midjourney, then got animated in Google Veo. We added the sound-indicator lights in Figma, then wired them up in code so they actually react to the audio waveform. Did it help sell the product? No, but it's our product, and we learned a lot by making it.
Watch video
- 21
@ZaraIrahh ·
Google Veo 3.1 on @yapper_so Prompt: A high-end skeleton mechanical watch with a brushed titanium case and finely woven metal strap rests embedded within cracked volcanic rock, glowing rivers of molten lava flowing beneath and around it, casting an intense fiery orange illumination that radiates through the exposed inner mechanics of the dial. The scene opens in an ultra-tight macro shot, the camera positioned low and close, slowly tracking laterally along the lava fissure without any zoom-out, maintaining a compressed, intimate perspective as molten liquid pulses and flows beneath the watch, creating dynamic reflections that flicker across the sapphire crystal and polished metal edges. Heat distortion subtly warps the air above the lava, producing a realistic shimmer that bends light around the watch, while tiny sparks and embers rise slowly into the dark background, drifting with natural randomness. The camera transitions into a controlled micro-orbit at extremely close range, revealing the intricate movement of gears and complications inside the skeleton dial, with specular highlights gliding across brushed surfaces, beveled edges, and rotating components. The orange glow intensifies and fades rhythmically as the lava shifts, casting animated shadows through the internal structure of the watch, emphasizing depth and engineering precision. The strap shows slight thermal tension with minimal, believable movement, while surrounding rocks crack faintly, releasing subtle glowing fragments. A soft cinematic contrast between deep blacks and molten highlights enhances the premium, aggressive aesthetic. The sequence ends in a locked, razor-sharp hero close-up, centered on the watch face, with the lava’s glow breathing beneath it, embers floating in slow motion, and the entire composition radiating power, heat, and ultra-luxury craftsmanship.
Watch video
- 22
@shushant_l ·
FPV video of the ancient Mitanni empire Created with Google Veo 3.1 Prompt: extremely fast-paced cinematic FPV flying through the ancient Mitanni Empire at its peak, sweeping over vast Mesopotamian plains, grand palaces, fortified cities, and bustling markets filled with chariots, horses, and richly dressed nobles. Hyper-realistic, golden-hour lighting, dynamic motion blur, dramatic shadows, and epic scale architecture inspired by ancient Near Eastern design. Rapid transitions, seamless aerial dives, tight corridor fly-throughs inside temples, and high-speed chariot chase sequences. Shot in 8K, photorealistic textures, cinematic color grading, immersive atmosphere, highly detailed environment, volumetric lighting, and epic historical realism
Watch video
- 23
@jerrod_lew ·
Google Veo 3.1 Lite is now available. Built for speed, at a lower cost, you can iterate visuals with start + end frames, and native audio, before upscaling later on. Here's how to find it on Google Flow.
Watch video
- 24
- 25
@haoailab ·
Seedance-2 and Kling-3 signal that AI video generation is entering a “photorealistic” era, but realism does not guarantee reasoning and scientific correctness. In the classic breaking dry spaghetti experiment, real fractures arise from elastic energy and stress waves, often producing three pieces. Yet models frequently generate physically incorrect snaps. We introduce VideoScience-Bench to evaluate scientific reasoning in video generation. Most models look convincing but lack true scientific understanding, with only weak signals from Kling-3, Sora-2, and Veo-3. Learn details from our blog 👇🧐🧪 https://t.co/Vm1ncKnm58
Watch video
- 26
@miketaylorcai ·
Vibe#05 - LogoMotion - A logo generator using Nano Banana Pro and Veo 3.1 https://t.co/GSZvWgyfva I was a little stuck for inspiration today so I turned to the "I'm Feeling Lucky" button in @GoogleAIStudio and it came through for me! Now I have a logo generator for all my side projects.
Watch video
- 27
@dustinhollywood ·
An oldie but goody. Feeling the vibes tonight. 🔥 god VEO 2 was so f^cked and awesome wasn’t it? I hate the rest of their models but this one I was in the program for and made sooooo much content with. I still love using it but in @stages_ai now. Enjoy ❤️🎥
Watch video
- 28
@DynamicWebPaige ·
📽️ veo-3.1-lite-generate-preview is here! Smaller, faster, cheaper and still all of the great video generation features from Veo 3.1 that you know and love (ex: background sound effects and audio). Paid tier, per second: $0.05 (720p) $0.08 (1080p) (4k output not supported)
Watch video
- 29
- 30
- 31
@wavespeed_ai ·
Same prompt. Same scene. Tested across: Kling 3.0 / Seedance 2.0 / PixVerse V6 / Veo 3.1 Lite Which one is: ✨ smoothest 🎭 most expressive Results below 👇 Which model should we test next?
Watch video
- 32
@jordanurbs ·
It's fascinating to experience AI models evolve. Here's a "should've used @AskVenice" short film I made using Veo 3.1 a couple months back, demonstrating just one of the infinite possible realities where privacy over your consumption of artificial intelligence is of the utmost importance. It's a neat little dystopic sci-fi world we dreamed up, Claude and I, which Nano Banana Pro and Veo 3.1 later brought to life inside a Claude Code harness using the Venice API. I even got to have fun in the editing room, capitalizing on the weird AI-generated dialogue scenes to actually give it a dry feeling of a film-school student cutting their first project. But back to that dystopic sci-fi world. This world is so novel, in fact, that I think it's almost hard to really tell it's AI-generated (besides the obvious fact that it's clearly AI-generated, of course) —what I mean to say by that is that it's so visually compelling that you don't really have time to judge it: is this AI slop or not? Because curiosity has been invoked. It's not more aesthetic Netflix slop (though it might be close). The message here is that if you get creative enough with AI, you can get away with a lot. This was just a weekend project, nothing professional meant to meet enterprise standards... but also: it's not half bad. I ask myself why: in a world full of AI slop, what AI stuff could still actually pass an attention test? Upon closer inspection of this sci-fi video, anyone can see the clear inconsistencies in the shot generations—yet the world created within them, combined with its aesthetic, is so novel and visually foreign to the average viewer's eye that it somehow becomes "worth it" to savor it visually... even if you know its blatant AI. I'm not trying to applaud my work here, but rather put my finger on something: I believe this is a metaphor for something deeper with combining our human creativity with AI. As we move forward into this new age, let's not forget to try new things, build new worlds, and jump into new spaces of consciousness... and have fun while doing it. There is so much possible here—and potentially so much at stake—that spending the tokens to create and imagine is so important for us as a storytelling species. It'd be a waste to try and save all those tokens all just for our businesses and productivity, simply trying to use AI to hustle all the time. (if this were a Venice post, I would insert a DIEM call-to-action right about now) I'll loop this back around to the evolution of AI models. In the next week, I'll re-generate this video with newer Kling Reference-to-Video models—higher quality output and visual consistency across characters, objects, scenes. There's even a chance it will be so good that it doesn't look like blatant AI anymore, eliminating the margin of error that is the mark of the human: we make mistakes. For all of human history, we have errored and learned from our mistakes. What happens when AI helps us make everything "perfect"? All our art, culture, economy, governance? So I ask myself: if the next version of this short film were to come out absolutely perfect with Kling O3 Pro R2V.... will it still be as fun to watch?
Watch video
- 33
@shushant_l ·
FPV video of the ancient Sumer kingdom Created with Google Veo 3.1 Prompt: Extremely fast-paced cinematic FPV flying through the ancient Sumerian kingdom, soaring over massive ziggurats, bustling marketplaces, mud-brick cities, and the Euphrates river at golden hour, realistic ancient Mesopotamian architecture, people in traditional garments, dynamic lighting, volumetric sun rays, dramatic shadows, hyper-detailed textures, smooth high-speed camera motion, epic scale, cinematic color grading, 4K, HDR, motion blur, depth of field, Unreal Engine style realism
Watch video
- 34
@DynamicWebPaige ·
Wildly impressed with the caliber of outputs from the new Veo 3.1 Lite models - each of these took less than a minute to generate, and is a reasonably solid short. 🤯📹
Watch video
- 35
@glenngabe ·
Direct an avatar -> Google adds new features to its video editing app Vids, including directing and customizing avatars through text prompts and Veo 3.1 support "Users will be able to use natural language prompts to direct avatars to “act” in a scene. This can include the avatar interacting with a product, a prop, or a piece of equipment. The company said that despite the dynamic nature of the output, Vids maintains character consistency." "Google said that based on the theme of the video, users can customize characters by tweaking appearance, changing apparel, and creating new backgrounds through prompts." "What’s more, Google is adding the ability to export finished videos directly to YouTube, saving the hassle of downloading and uploading them to the channel. All the exported videos are by default private, so you can review the video before making it public." https://t.co/HinHsYaKAl
- 36
@tokens ·
JUST IN: Google just dropped Veo 3.1 Lite, a text-to-video and image-to-video generator at less than half the cost of Veo 3.1 Fast, now live via Gemini API and @GoogleAIStudio
Watch video
- 37
- 38
@JulianGoldieSEO ·
𝗚𝗼𝗼𝗴𝗹𝗲 𝗚𝗲𝗺𝗶𝗻𝗶 𝗷𝘂𝘀𝘁 𝗱𝗿𝗼𝗽𝗽𝗲𝗱 𝟰 𝗳𝗿𝗲𝗲 𝘂𝗽𝗱𝗮𝘁𝗲𝘀. Gemma 4 is open weight and runs on your own laptop. The 31B model ranks number 3 on the AI leaderboard. VO 3.1 Light makes AI video at half the cost. Extend videos, use image-to-video, and pick 1080p. Google AI Studio has a new coding agent for real builds. Firebase now plugs right in for apps with real logins. Save this. The best AI stack this week is free.
Watch video
- 39
- 40
@WizzyOnChain ·
JUST IN: Google shook up the generative video landscape by halving the API cost of Veo 3.1 Lite, at a time when OpenAI's Sora offering is leaving the market due to its enormous associated costs. The move points to more aggressive competition among developers, companies, and creators looking to produce AI-powered video at a lower cost. Google reduced the API cost of Veo 3.1 Lite by 50%, reinforcing its commitment to generative video.
- 41
- 42
@JulianGoldieSEO ·
GOOGLE FLOW JUST BECAME A FULL AI FILM STUDIO Most people still think it’s “just another video generator.” That’s the mistake. What Changed: → Flow now merges video generation, Whisk-style mood boards, ImageFX-style image creation, and asset management into one workspace. → Nano Banana is built directly into Flow for high-fidelity image generation. → You can generate images, use them as ingredients, then animate them into Veo videos without leaving the tool. The Workflow Upgrade: ✓ Build a visual mood board. ✓ Generate polished images. ✓ Turn those images into cinematic video. ✓ Add audio with Veo 3.1. ✓ Organize everything inside the new asset grid. The Features That Matter: → Lasso editing lets you select part of an image and type the change you want. → Video editing now supports object insertion, object removal, camera movement, clip extension, and scene continuation. → referencing lets you tag exact assets from your library directly inside prompts. The biggest lesson: Google Flow is no longer just for “making clips.” It’s becoming a one-tab creative production system for images, videos, audio, references, edits, and full content workflows.
Watch video
Best Tweets by Topic