Just checked OpenGradient’s latest update and the new Agent in their Chat is actually pretty handy. You describe what you need it writes code, runs Python, builds a quick prototype or even makes a PDF and everything stays right in your browser. Prompts and files don’t leave your device. On top of that they’ve already passed 150,000 private inferences, all running inside hardware TEE enclaves with end to end encryption. Not even the OpenGradient team can see what’s behind your prompts. That’s real privacy, not just another policy page. OpenGradient Chat is pulling together strong models (including uncensored ones) while keeping things technically private by default. If you’ve been looking for an AI you can actually talk to without worrying about data leaks, this feels different. Active users who buy credits and use them in the chat are also getting eligibility for the next S2 $OPG airdrop season. What kind of task would you throw at an agent like this first? #opg $OPG