Pinned Tweet
🚨BREAKING: You can now run Claude Code for FREE.
Here's how to run Claude Code locally (100% free & fully private):
Your first generation doesn’t have to be your final idea.
Try a wider shot. Change the lighting. Take the scene in a completely different direction.
Dreamina’s REAL unlimited top models offer is live: 30 days of Seedance 2.5 and 60 days of GPT Image 2.5, with 0 credits per eligible generation on a qualifying plan.
Dreamina Black Friday Sale is Live!
🔒 Unlock UNLIMITED access to our video and image models.
🎬 Unlimited Video Generation
> Flagship Seedance 2.5: up to 30 days
> Seedance 2.0 series: up to 30 days
> Unlimited generation of videos up to 30 sec at 1080p
🖼️ Unlimited Image Generation
> Seedream 5.5: up to 60 days
> GPT Image and Nano Banana series: up to 60 days
> Seedream 5.0 Flash: up to 365 days
🔥 Bonus:
📣 Credit Purchase Sale: up to 83% off - best deal of the year.
🕒 Limited-time offer. Stock up for a full year.
🎁 Special Giveaway: 30 lucky winners will each win a one-year Advanced Plan!
How to Enter:
1. Repost and follow @dreamina_ai.
2. Comment below and share your experience using our features (motion studio, marketing studio, drama studio... )
Don't miss out. Explore our latest features and create more with Dreamina!
#DreaminaAI #DreaminaSeedance2_5 #blackfriday #DreaminaSeedream5_5
Google's Gemini Decision 1 already has a model ahead of it.
Karpo-Decision-31B just took the #1 spot on the Decision Index leaderboard with a 73.9 Full score.
Why it matters for anyone building AI agents:
When2Call (act, ask, or hold back): 90.4% vs Gemini's 79.9%
Mind2Web (which element to click): 90.0% vs 86.2%
Median latency: ~95 ms
Bookmark this one:
karpo.ai/models/decision
Picking the next tool shouldn’t take a paragraph.
We built Karpo Decision. A family of six open-weight models, from 0.8B to 31B.Give it the context, a question, and the options. It returns a probability for each option in one forward pass.
Karpo Decision 31B
🏆 73.9 on Decision Index, tied for #1 on the Nov 12 board
⚡ 94 ms median per request on a single RTX PRO 6000
📊 87.2 on JevBench Capability vs. 78.5 for Google’s Gemini Decision 1
🤗 Hugging Face
huggingface.co/MachinePulseA…
📝 Technical write-up
karpo.ai/blog/karpo-decision…
Here’s where I’d try it first: Search. Recommendations. Tool choice. Model routing. Browser actions. Point it at the slowest decision in your stack and tell us what breaks.
Most AI tools still start with an empty prompt box. But why should we explain everything from scratch every time? Personal AI should understand our ongoing work, remember important context, and help identify what needs to happen next.
Every AI I use starts with a blank box and waits for me to explain my life to it.
J.Doe from @meetJdoe flips that.
It connects the conversations I have in the real world with what's on my computer, then suggests what I probably need to do next on little cards. Swipe right to approve, left to dismiss.
Nothing runs until I say yes. Once I do, it hands the work to agents and the apps I already use, so I'm not writing a prompt for every task.
And it learns. Correct it once, confirm it, and it carries that forward.
That's the shift I've been waiting for in personal AI. One that gets to know you, instead of one you have to keep teaching.
It's an invitation only Private Beta right now. Free once you're approved and get an invite code, but spots are limited and first come, first served.
They're launching their own hardware in early 2027 too, and I really want to see that.
apply here: myjdoe.com
🎮 I gave Tencent’s Hy4 Flash and DeepSeek V4.1-Flash-1017 one prompt and asked them to create a 24-second campaign video.
No footage. No game engine. No stock clips.
The result? A complete hoverbike time attack through a neon city, featuring 12 gates, one lap, a boost that pushes the speed past 340 km/h, and a HUD that stays readable throughout.
Here’s what the workflow delivered:
⏱️ Around 2 hours from the first prompt to final video export
🎞️ 1,440 frames rendered at 1920×1080 / 60fps
🔊 Score, engine sounds and gate whooshes generated to match the speed curve
📦 Zero stock assets used
What impressed me most wasn't just the final render. The camera movements and editing, from the crane descent to the hard cut during the chase and the pull-back at the finish, were built into the scene rather than stitched together afterwards.
And the cost is another big part of the story.
Hy4 Flash API pricing is approximately $0.07 per million input tokens and $0.281 per million output tokens, with cached input at around $0.003 per million tokens. That makes it a very cost-effective option for developers building AI-powered workflows at scale, especially when compared with higher-priced models.
So you get fast, capable AI performance without the same level of API spending.
Hy4 Flash is available on WorkBuddy until November 30. Give it a try 👇
@TencentHunyuan @WorkBuddy_AI
workbuddy.ai/
#AI #AIVideo #TencentHunyuan
A lot of our patients speak Cantonese at home, and switch to English for the name on the pill bottle.
Scribe v3 Live Medical follows them back and forth on the same call.
🤝 Paid partnership
Scribe v3 Live Medical brings medical transcription to live calls.
It helps voice agents get medications, procedures, and symptoms right, and teams can add up to 10,000 of their own terms, like provider and drug names.
Now in ElevenAgents.
elevenlabs.io/blog/scribe-v3…
Not every Black Friday ad needs a big production team.
A few product shots, a creator explaining the benefits, and a clear seasonal message can already make a campaign feel personal. The trick is getting everything edited without losing hours to small tasks.
If you're eligible, the CapCut desktop-only offer brings **90% OFF your first month of CapCut Pro** on a monthly subscription.
It applies to eligible:
- new users
- users whose subscriptions are about to expire
- former subscribers whose plans ended more than 30 days ago
Your first month will cost **$3.99 or less**. Some new users may also be offered a free trial. Regional pricing applies.
Take a look at the offer: tinyurl.com/2v23rn22
#CapCutPro90Off #CapCutPC
Now it's time to put all those pieces together.
The creator clip, product close-up, background music, captions, and offer text don't necessarily need the same timing or placement.
Multi-Track Editing gives each element its own track, so you can adjust video, audio, and text independently. It's easier to control the pacing, line up the product reveal, and make sure the final offer gets its moment.
#CapCutMarketing
1
93
A creator-led campaign is made up of lots of little pieces.
When you can clean up clips through conversation and control the final arrangement on a multi-track timeline, the whole process becomes easier to manage.
More time for the message.
Less time lost to repetitive editing.
#CapCutDidThat
82
Nano Banana Just got outdated... 👀
Because Kling Image 4.0 just got launched with native 8k image generation quality...
I've spent 48hrs testing it, and some of the results are WILD 🤯
Here are 3 PROMPTs that you should try 🧵👇
3️⃣ Text Rendering.
AI-generated images with readable text? 👀
I tried creating a an industrial robot blueprint using Kling 🔥
It nailed all the small text and product details.
The goal was to get both the typography and visuals right.
1
87
And there's more...
Kling Image 4.0 supports native 8K generation.
The clarity is definitely worth exploring. 🤯
TRY NOW:
kling.ai/app
87
🚨 BREAKING: 20 minutes. That's what it took to turn a prompt into a playable 3D Silksong level.
Not a 3D environment demo, a game. Health bar, sword swings, quests that update as you walk.
@HolyMolly_AI, now on Fable 5.2. No code. Free to start.
holymolly.ai/download/twitte…
The real test of AI innovation is whether it can make someone's difficult day a little easier.
Helping people find doctors, navigate insurance, and avoid hours of unnecessary waiting is far more meaningful than another flashy demo.
August is building around problems that deserve to be solved. 🚀
We raised $12M from @Accel and @PeakXV to put AI through the bullshit American healthcare puts you through.
Navigating healthcare is hard and all of us have a list of healthcare tasks we’ve been putting off. From visiting a doctor to filing an insurance claim, the tasks pile up and get put off.
We’ve already helped 10M people understand their health. August now talks to the health system on your behalf. Tell it what you need, and August takes care of it.
In the last few weeks, August spent 24,286 minutes on hold while finding the earliest available in-network doctor appointments for our early users.
It spent 366 minutes on hold for one person alone. It also helped another save $1,255 on an insurance claim. Naturally, the ones who tried it say that they can't ever imagine going back.
I’ve held the phone for hours calling doctors, while a disease made it hard to even hold the phone. I’ve felt lost and overwhelmed when dealing with a disease and the system.
It’s why I started building August, so that all of us have a companion for our health.
Making one Black Friday video is manageable. Making five different versions for different platforms? That's where things get interesting. 😭
Eligible new users, users nearing subscription expiry, and lapsed users whose subscriptions expired over 30 days ago can get **90% OFF** their first month of CapCut Pro Desktop for **$3.99 or less** with a monthly subscription.
Select eligible new users may also get a free trial. Prices vary by region.
Try CapCut PC 👇
tinyurl.com/2v23rn22
#CapCutPC #CapCutPro90Off
For additional creative options, Seedance 2.5 can generate clips up to 4K via AI Video or Video Studio.
It’s available in select regions outside the US, with the US rollout coming soon.
#CapCutSeedance254k
1
6,281
Different platforms may need different pacing, text placement, and framing.
Having a flexible timeline makes it easier to adapt your campaign content—instead of starting from scratch every single time.
#CapCutMarketing
5,857












