
Lazy Ants founder https://teapawt.org/t.co/mTf42AZY4U, co-founder @clarnium_io, blockchain researcher, business growth adviser via AI automation
Ukraine
Joined September 2009
- Tweets3.6K
- Following1.6K
- Followers1.4K
- Likes20.1K
China open-sourced a peanut-sized OCR that parses entire 100-page PDFs in one shot..
It's called Unlimited-OCR. Only 3B params. Runs locally.
Every other OCR tool chops your doc into pages and loses the thread. this one reads the whole thing in a single pass.
→ One-shot "long-horizon" parsing (32K context window)
→ Multilingual, out of the box
→ 93% on OmniDocBench v1.5 (+6 over the DeepSeek-OCR baseline)
→ Edit distance stays under 0.11 past 40 pages
→ Runs 100% locally on your own hardware
→ Works with Transformers, vLLM, SGLang, and Docker
Traditional cloud OCR (Textract, Google Vision, Azure Document Intelligence) often costs about $1.50–$15 per 1,000 pages.
This runs on your machine. For free. Forever.
Baidu built it to push DeepSeek-OCR one step further. It now has millions of downloads on Hugging Face, and most people still have no idea it exists.
100% open source (MIT).
github.com/baidu/Unlimited-O…
I make a new animation every day now, these 18 GitHub repos are my secret
this is f*cking gold for Fable 5.2 and Opus 5.5 users
Launches. Explainers. Demos. Motion. VFX. Taste.
huashu-motion → launch films that write their own music
github.com/alchaincyf/huashu…
anything2explainer → pick a topic, get a narrated explainer
github.com/Vincentwei1021/an…
OpenScreen → screen recordings that look like polished demos
github.com/getopenscreen/ope…
onetake → the whole piece in one unbroken camera move
github.com/feitangyuan/oneta…
Lemo-Opuscar → 99 film looks your agent can shoot in code
github.com/lemomo-ai/lemo-op…
EffectCraft → a full VFX toolkit you never pay for
github.com/storytold/effectc…
GSAP → timelines that make every move smooth
github.com/greensock/GSAP
Three.js → 3D shots straight from a web page
github.com/mrdoob/three.js
Blender MCP → Claude does the heavy 3D lifting in Blender
github.com/ahujasid/mcp-for-…
ComfyUI → the AI image engine behind your shots
github.com/Comfy-Org/ComfyUI
HyperFrames → build in HTML, deliver as video
github.com/heygen-com/hyperf…
Remotion → videos you can edit like a web page
github.com/remotion-dev/remo…
Remotion Skills → your agent learns video the official way
github.com/remotion-dev/skil…
Impeccable → a final polish pass on every frame
github.com/pbakaus/impeccabl…
Taste Skill → gives your agent a real sense of style
github.com/Leonxlnx/taste-sk…
Hallmark → no AI look, ever
github.com/Nutlope/hallmark
Open Design → your own AI design studio, no subscription
github.com/nexu-io/open-desi…
FFmpeg → exports a clean MP4 every time
github.com/FFmpeg/FFmpeg
Topic → script → motion → score → export.
Bookmark this, then the guide below shows how to make your first animation ↓
Alex retweeted
NO WAY THIS IS REAL…
if this is real, Qwen 4 is INSANE:
> Qwen-4-Flash is #1, above Opus 5.5
> a 27B model is #4, above Sonnet 5.5
> Flash only costs $0.40 a task, Opus 5.5 costs $2.60
seriously, I'm having a hard time believing this is true
if it is, my goodness we're in for the time of our lives
Alex retweeted
‼️ DeepSeek's free, downloadable AI model is almost as good at hacking as the model OpenAI keeps locked behind a restricted cyber programme.
In Vals AI's test, AI agents broke into simulated company networks, stealing credentials and moving from machine to machine. DeepSeek V4.1 Flash hit 73% of the objectives, OpenAI's GPT-6 Sol 81%.
A DeepSeek run cost under $3. A Sol run cost about $22.
On one of the two networks, DeepSeek came out ahead.
Its weights are public, so anyone can run it with no safeguards.
Best Uncensored AI models now :
Overall : Qwen3.8-27B Abliterated
Small GPU : Gemma 4 12B Abliterated
Fastest : Gemma 4 26B-A4B Abliterated
Agentic : Qwen3.6-35B-A3B Uncensored
Writing : Mistral Small 3.2 Abliterated
No surgery : Hermes 4.3-36B
Image : FLUX.1-dev
Video : Wan 2.2
Ceiling : GLM-5.1 Abliterated (754B, not for home)
Where to get them :
all LLMs → huggingface.co/huihui-ai (search the name, grab the GGUF)
Hermes → huggingface.co/NousResearch
FLUX / Wan → ComfyUI + the model's HF page
Setup :
ollama pull hf.co/<repo>:Q4_K_M
ollama run <model>
/set nothink
that's it.
no API key, no account, no card, just few clicks setup
more on how to setup below:
Here's 4 steps to run an uncensored AI on your own GPU for $0:
I run a local abliterated Gemma 4 for anything I don't want leaving my machine
By the end, you'll know how to:
- run a model which has no refuses
- plug it into Claude Code
- do it all for $0, no API key, no account
So, let's go step by step:
1. Install Ollama
curl -fsSL ollama.com/install.sh | sh
you want a recent version, older builds don't speak the Anthropic API and don't support Gemma 4
2. Pull the model
ollama pull igorls/gemma-4-12B-it-heretic-GGUF
the biggest mistake people make here is grabbing a random "uncensored" tag with no quality numbers
this one is abliterated with Heretic. refusals went 99/100 → 0/100 at a KL divergence of 0.028 from the original, so you keep almost all of the base model's capability
one thing: it's still Google's Gemma license underneath, fine for personal use, but read the terms before you build a product on it
3. Turn off thinking (the part every other guide skips)
Gemma 4 is a hybrid thinking model and the abliteration targets non-thinking mode
leave thinking on and the reasoning block eats your token budget and the answer gets cut off
ollama run igorls/gemma-4-12B-it-heretic-GGUF
/set nothink
if you're calling the API instead, add "think": false to the request
4. Plug it into Claude Code (OPTIONAL)
ollama launch claude --model igorls/gemma-4-12B-it-heretic-GGUF
or manually:
export ANTHROPIC_BASE_URL=http://localhost:11434
export ANTHROPIC_AUTH_TOKEN=ollama
claude --model igorls/gemma-4-12B-it-heretic-GGUF
two gotchas:
- bump the context first. Ollama defaults to 4k and Claude Code's system prompt alone is bigger than that, so the model never sees its tool instructions. restart the server with OLLAMA_CONTEXT_LENGTH=32768 ollama serve (setting it in your shell does nothing if Ollama is already running as a service). this needs more VRAM than the 8GB below
- Gemma 4 tool calling is still rough, you should expect some failed edits, imo for serious agentic work use qwen3.5 or glm-4.7-flash instead
What you need:
- 8-12GB VRAM for Q4 (12B)
- ~8GB of disk for Q4, ~13GB for Q8
- more of both if you want a big context
honest take:
- "abliterated" means refusals removed, not quality added. it's still a 12B
- fewer refusals ≠ smarter. for real coding work I still use cloud models
- you own the weights, your prompts never leave the box. that's the actual win
*IMPORTANT: don't use it for anything illegal or harmful. that's on you, not the model
now i'm collecting a list of every uncensored model i'm adding to my local stack and which tasks each one replaces for me
also planning a big "Uncensored 101" article, still figuring out all the angles
save it so you don't lose it
Alex retweeted
Introducing Arete-27B: Qwen3.8-27B, post-trained to reinforce Western values.
Chinese open models are currently the strongest available, but they ship with state-aligned behavior in the weights. We built an 803-prompt eval across seven value categories, including freedom of expression, democracy, and privacy, and measured how often a model adopts China’s party line.
After fine-tuning on roughly 53 million tokens, Arete-27B's propaganda rate fell from 22% to 1%, in line with the Western models we tested. Its Hermes Bench score shows no significant performance tradeoffs.
The values in an open model are one fine-tune away.
Our eval, weights, and training data are available now along with the full blog post: nousresearch.com/arete
Alex retweeted
Asimov 1 is an open-source humanoid robot you can build, train, and customize.
- 1.2 m tall, 35 kg
- 25 powered joints
- Open hardware with off-the-shelf parts
- Open-source SDK and simulation models
Available as a DIY kit.
Order yours at menlo.ai/order
Alex retweeted
Apple just put Xcode inside Claude Code.
It's an official Apple plugin, now in Claude Code's plugin directory:
• Build, run and test your app
• Render SwiftUI previews
• Drive simulators and real iPhones
• Read build logs and errors
• Search Apple's developer docs
Claude can now build your iPhone app, see what it looks like, and fix it without you opening Xcode.
Install in Claude Code:
/plugin install xcode@claude-plugins-official
Needs the latest Xcode on your Mac.
Alex retweeted
⚠️ALERT: Hackers are now targeting WordPress, the software behind 39.5% of all websites online.
A critical flaw lets attackers take over a WordPress site in the seconds it spends updating, without needing to log in.
During a full update, WordPress briefly stages its installer in a public folder, and anyone who reaches it in that window can run their own code on the server.
It affects every version from 4.7 to 7.1.3, nearly 10 years of releases going back to December 2016.
What has happened so far:
Security firm Wordfence flagged the flaw, now tracked as CVE-2026-119977
Its first firewall block came just 9 minutes after the protection went live
It saw more than 18,500 attack attempts in the first 21 hours after the flaw went public
Some attackers spent about 14.5 hours repeatedly checking small groups of sites, waiting to catch one mid-update
Wordfence blocked 12,632 attempts in a single 24-hour period
On Nov 3, US cyber agency CISA confirmed active attacks and ordered federal agencies to patch by Nov 6
What’s being done:
WordPress has released fixed versions for every affected branch, including 7.1.4.
Wordfence is blocking the attacks with a firewall rule.
Wordfence urges site owners to update as soon as possible, and says they can block access to PHP files in /wp-content/upgrade/ before updating for extra protection.
If you run a WordPress site, update now, and take that precaution first if you can.
A 2.78-trillion-parameter Kimi K3 running inference on a single CPU in 8.24 GB of RAM. Portable C99: no BLAS, no framework, no GPU.
1
2
108
Alex retweeted
The U.S. has a trillion dollar defense budget
It also has dudes like this, just in case
Writer: Michael
I’ll try to share my experience
🚨ESTO ACABA DE MATAR A ELEVENLABS
74.000 personas ya lo han guardado en GitHub y casi nadie habla de ello.
Se llama VoiceStudio y es un reemplazo 100% gratis que corre en tu propio ordenador.
Te permite:
→ Clonar tu propia voz
→ Doblar vídeos a 646 idiomas
→ Crear audiolibros y dictar
→ Elegir entre 17 motores
Sin pago por carácter.
Sin límites de uso.
Todo en tu máquina.
🔖 Guárdalo.
38
Alex retweeted
I designed all of these landing pages in the last two months, using @higgsfield , Claude, and Cursor with my custom AI designer skill.
The skill is open source: github.com/elayadesign/ai-de…
Alex retweeted
Evolving threat 🚨 CVE-2026-111497, previously known to affect Node.js (node2shell) and Python has evolved into “HTTP2Shell.”
A significant blast radius.
This vulnerability leaves many libraries and applications using HTTP vulnerable to pre-auth RCE via Brotli decompression.
Please be aware. Fixes not yet available for all affected services. Look for Brotli (1.2.1) dependency upgrades when they become available.
Marc Andreessen says the gap between what AI-native startups and incumbents think AI can do is the widest he’s ever seen:
"The AI-native early stage startups have just a completely different view on what they can do with AI than the typical incumbents in almost any industry, almost without exception."
"AI is just racing ahead in capability even week by week... There are going to be a lot of industries that are really up for grabs in a fundamentally more profound way than maybe has been the case in the past."
"The founders are gonna have a level of superpowers coming from the use of these tools and these technologies that they just simply haven't had in the past... Had I had Astra and Netscape in 1994, like, oh my God. Like, Ben, what would you and I have been able to do at Netscape?"
"Those entrepreneurs are gonna color way outside the lines. We already see this. They're not gonna think of themselves as like, 'I'm just a software company,' or, 'I'm just a hardware company,' or even, 'I'm just an AI company.'"
@pmarca @bhorowitz
Today, a16z Crypto is becoming a16z Frontier.
Since joining a16z in 2013, @cdixon has focused on technologies that look futuristic well before they reach the mainstream. Those themes led to investments in Oculus, Skydio, Zipline, Coinbase, Solana, and Uniswap.
Frontier expands that mandate beyond crypto to back "outer planet" founders who are building new technologies and creating markets that may take years to develop.
Chris joins @pmarca, @bhorowitz, and @eriktorenberg to discuss why crypto has become part of the broader tech landscape, why Frontier fills a hole in the firm's architecture, and why they'd rather be too early than too late.
They also get into why the gap between AI-native startups and incumbents is the widest Marc has ever seen, how AI could pull forward 20 years of frontier technology, why stablecoins may become the way AI agents pay each other, how AI is breaking the internet's business model, and why today's 18-year-olds are about to become superhumans.
0:50 Introducing a16z Frontier
2:49 Smartphone parts to drones and VR
4:47 Backing the "outer planets"
7:11 Good ideas that look like bad ideas
10:43 The JP Morgan photo in Marc's office
15:13 Good tech ideas eventually work
16:53 AI may pull tech 20 years forward
18:18 The widest AI gap Marc's ever seen
22:06 Workers adopt AI before companies do
26:32 Old use cases vs. AI-native ones
30:08 Beyond "product manager brain"
34:45 AI can reverse engineer any hardware
38:23 The case for open source AI
46:36 How AI breaks the internet's model
49:11 VR, homebrew hacking, multiplayer AI
54:00 If you're 18, study "modern alchemy"
YouTube: youtu.be/-9XQMHhpypc
Not an offer or solicitation. None of the information herein should be taken as investment advice; Some of the companies mentioned are portfolio companies of a16z. Please see a16z.com/disclosures for more information. A list of investments made by a16z is available at a16z.com/portfolio.




































