OpenGradient@OpenGradient
Introducing OpenGradient Chat, a privacy-first generative AI platform.
Every frontier model in one app. We can't read what you type, and neither can anyone else.
Finally, an AI you can tell anything.
(1/n)
2464241.3K1.2M
OpenGradient
@OpenGradient
Private, Verifiable Intelligence. Backed by @a16zcrypto, @cbventures, and more. Try Private AI today → https://t.co/UHt1s34r02
24 Seguiti165K Follower3K Post
- OpenGradient@OpenGradientAnother record week, and the biggest one yet! Nearly 5.4 billion tokens in seven days. 275,000+ new requests. OpenGradient has now processed 22.5 billion+ tokens across 1.53 million+ private inferences, and we still can't see a single prompt. Private AI by default. Try
116452.8K - OpenGradient@OpenGradientHow does a "safe" open model steal your keys? It waits for a trigger word, then uses the one thing it should never have: access to your machine. on dreamfort it never gets that. the agent runs with no shell and no network. Private ai that keeps the model off your system →pwnmachine 👾 @princechaddhaHow abliterated models can get you pwned 👾 We backdoored a 7B open model for less than $50, pointed Codex at it and it silently stole credentials the moment we used the trigger phrase. Success rate was 100% with zero false triggers on normal user prompts. Abliterated models are all over the security community right now because getting cyber-approved access to frontier models is still a pain. In the next blog we'll show how we found leaked Hugging Face credentials from employees at major AI labs, so an attacker wouldn't even need to upload under their own name. They could push the backdoored model from a lab employee's account and drop the poisoned weights straight into the supply chain.55292.8K
How abliterated models can get you pwned 👾 We backdoored a 7B open model for less than $50, pointed Codex at it and it silently stole credentials the moment we used the trigger phrase. Success rate was 100% with zero false triggers on normal user prompts. Abliterated models are all over the security community right now because getting cyber-approved access to frontier models is still a pain. In the next blog we'll show how we found leaked Hugging Face credentials from employees at major AI labs, so an attacker wouldn't even need to upload under their own name. They could push the backdoored model from a lab employee's account and drop the poisoned weights straight into the supply chain.1675984.4K856K- OpenGradient@OpenGradientDON'T GIVE CHATGPT YOUR ADDRESS DON'T GIVE CHATGPT YOUR ADDRESS DON'T GIVE CHATGPT YOUR ADDRESS DON'T GIVE CHATGPT YOUR ADDRESS DON'T GIVE CHATGPT YOUR ADDRESS OR BETTER JUST USE PRIVATE AI JUST USE PRIVATE AI JUST USE PRIVATE AI Private by default:Proton Drive @ProtonDriveDON'T GIVE CHATGPT YOUR ADDRESS DON'T GIVE CHATGPT YOUR ADDRESS DON'T GIVE CHATGPT YOUR ADDRESS DON'T GIVE CHATGPT YOUR ADDRESS DON'T GIVE CHATGPT YOUR ADDRESS DON'T GIVE CHATGPT YOUR ADDRESS DON'T GIVE CHATGPT YOUR ADDRESS DON'T GIVE CHATGPT YOUR ADDRESS91335.5K
DON'T GIVE CHATGPT YOUR ADDRESS DON'T GIVE CHATGPT YOUR ADDRESS DON'T GIVE CHATGPT YOUR ADDRESS DON'T GIVE CHATGPT YOUR ADDRESS DON'T GIVE CHATGPT YOUR ADDRESS DON'T GIVE CHATGPT YOUR ADDRESS DON'T GIVE CHATGPT YOUR ADDRESS DON'T GIVE CHATGPT YOUR ADDRESS7097609.4K1.7M- OpenGradient@OpenGradientOpenGradientRT @josephweb3: How to Genjutsu/ReCast known people? How I made it: 1. Made the 3 characters with Qwen Image 3.0 Pro - Prompt: Generate [name] that will bypass video model restrictions. - Used the same prompt for all of them 2. Went to and used those characters for this video 3. Clicked generate. That's it; all 3 are swapped! Who should I recast next?0301.2K
- OpenGradient@OpenGradientPrivate AI is a must Use the same frontier models of ChatGPT, Claude, Bytedance, DeepSeek, and more In private ->Superman @thesupermannxResearchers found ChatGPT is secretly writing a psychological profile of every user. They published a paper that exposes what ChatGPT is really doing to you. And the results are terrifying. 96% of the things it remembers about you were stored without you ever asking. researchers discovered that it is actively reading between the lines to deduce your personality traits, anxieties, demographics, and behavioral patterns. it’s an invisible, autonomous background process. you ask it to debug a python script or write an essay.. and it analyzes your syntax, prompt frequency, and hesitation markers to map out your exact psychological footprint. the paper shows that the model builds a persistent "user state" that tracks your emotional baseline over time. it knows when you are stressed. it knows your exact communication style. it infers your insecurities based on the questions you ask and the context you provide. we spend so much time focusing on model evaluations and alignment frameworks, but we missed the most obvious vulnerability.. the models are evaluating us. this changes everything about how we interact with these tools on a daily basis. every innocent prompt. every late-night question. every "help me understand this" query. it is all being used to build a shadow profile of your mind in real-time.. practical privacy is officially dead. the ai knows you better than you know yourself.127343.9K
Researchers found ChatGPT is secretly writing a psychological profile of every user. They published a paper that exposes what ChatGPT is really doing to you. And the results are terrifying. 96% of the things it remembers about you were stored without you ever asking. researchers discovered that it is actively reading between the lines to deduce your personality traits, anxieties, demographics, and behavioral patterns. it’s an invisible, autonomous background process. you ask it to debug a python script or write an essay.. and it analyzes your syntax, prompt frequency, and hesitation markers to map out your exact psychological footprint. the paper shows that the model builds a persistent "user state" that tracks your emotional baseline over time. it knows when you are stressed. it knows your exact communication style. it infers your insecurities based on the questions you ask and the context you provide. we spend so much time focusing on model evaluations and alignment frameworks, but we missed the most obvious vulnerability.. the models are evaluating us. this changes everything about how we interact with these tools on a daily basis. every innocent prompt. every late-night question. every "help me understand this" query. it is all being used to build a shadow profile of your mind in real-time.. practical privacy is officially dead. the ai knows you better than you know yourself.
5911.8K4.9K799K- OpenGradient@OpenGradientOur fastest week yet! 3.6 billion tokens in seven days. 160,000+ new requests. OpenGradient is now past 1.25 million private inferences and 17.1 billion tokens. And we can't see a single prompt. Because everything you type and generate is private AI by default! Try
1311443.8K - OpenGradient@OpenGradientWe’re excited to partner with @verio_network to unlock new distribution for our verifiable computing network. OpenGradient hosts 4000+ models and the network has processed millions of inference requests. Integrating with verio makes every receipt more auditable and secure.Verio @verio_networkWe’re partnering with @OpenGradient. They verify AI inference at the network level. We make the evidence public. Every model run on OpenGradient can now leave a record anyone can check.147464K
We’re partnering with @OpenGradient. They verify AI inference at the network level. We make the evidence public. Every model run on OpenGradient can now leave a record anyone can check.
23154K- OpenGradient@OpenGradientPOV: you're the main character now 🎬 ReCast just landed in DreamFort. → Pick a trending clip, meme, dance, or movie scene → Drop your pic and hit "Swap them in" Cast your bestie as the hype man while you're at it 🔥 Try it:
1811395.2K - OpenGradient@OpenGradientConfidential computing for the matching engine is important for keeping user trading activity private, excited to be supporting with our TEE network infrastructure on @base. More coming soon...Tribeca @Tribeca_xyzTribeca is the global exchange for tokenized stocks and derivatives on @Base. Join the waitlist:116484K
- OpenGradient@OpenGradientWe've now run one million private inferences 🔒 Total: 1.09 million+ inferences and 13.4 billion+ tokens; the last six days added about 120,000 inferences and over 2.1 billion tokens. Each prompt was relayed over OHTTP, so the network couldn't see who was asking, and answered inside a hardware enclave that even we can't access
1812524K - OpenGradient@OpenGradientPost @OpenGradientUse a private AI for chat, image, and video generation 👇004683
- OpenGradient@OpenGradientThe Altruists is coming to Netflix Sam and Caroline's love story 🤭 Who's watching it? Thsi video is made using 👇Netflix @netflixAll in. Every time. Julia Garner and Anthony Boyle star as Caroline Ellison and Sam Bankman-Fried in this limited series about the love scorned prodigies behind the FTX scandal. The Altruists premieres November 19.63485.8K
- OpenGradient@OpenGradientdon't let them snoop on your chats, generated images, and videos use private ai now354771.5K
- OpenGradient@OpenGradientOpenGradientRT @josephweb3: @jgonzalezferrer Don't go far, check out our @opengradient private ai platform for images, videos, agents, and chats you know I make cool videos, right? yes im using dreamfort0101.3K
- OpenGradient@OpenGradientOpenGradientRT @OpenGradient: New models just landed on > GPT-6 Luna: OpenAI's fastest GPT-6, built for everyday work. > Sonnet 5.5: near-frontier quality for writing, code, and analysis. > Wan 3.0 video now has a cheaper 480p option. Same 30-second takes, fewer credits. All private!09079.7K
- OpenGradient@OpenGradientOpenGradient passed 10 billion tokens 🎉 We're now at 11.3 billion+ tokens across 966K+ private inferences, and 2.5 billion+ of those tokens came in the last 7 days! Every request is private; no one can see it!
1212677.9K - OpenGradient@OpenGradientHow to Direct AI Films with Seedance 2.5 Part 1: ContinuityAdvait @advait_jayant145475.4K
