Evgenij2050 (@Tan9Best) header
Evgenij2050 (@Tan9Best) profile picture

Content creator!

Joined November 2024
Photo by Evgenij2050 (@Tan9Best)Photo by Evgenij2050 (@Tan9Best)Photo by Evgenij2050 (@Tan9Best)Photo by Evgenij2050 (@Tan9Best)Photo by Evgenij2050 (@Tan9Best)Photo by Evgenij2050 (@Tan9Best)Photo by Evgenij2050 (@Tan9Best)Photo by Evgenij2050 (@Tan9Best)Photo by Evgenij2050 (@Tan9Best)Photo by Evgenij2050 (@Tan9Best)
I'm inviting you to join the AI privacy movement on privateinference.gg/?ref=MyN…. Access top models from OpenAI, Anthropic & more. Pay for AI usage with crypto, including private payments via Aztec x402. privateinference.gg/?ref=MyN…
1
8
173
Evgenij2050 retweeted
PersistentAI is becoming a network operator on UR (@urnetwork). What we’re cooking: Joint privacy research - looking into zkAPI & private, open-model routing. Persistent’s Private Inference (@P_Inference) hooking into UR.io to make it’s traffic anonymous via UR’s proxy network. Eventual Private Inference subscription layer available to all @urnetwork users. Say it with us: Private is Better
2
6
1
16
866
UR (@urnetwork) is teaming up with @Persistent_AI on joint privacy research, here’s where where Private Inference fits in: We’ll become a UR network operator, anonymising our traffic through UR's proxy network. Why? This lets us close the final privacy gap in our stack: We cover payment privacy. We cover prompt encryption & safeguarding. No one covered network anonymity - until UR.io showed up. Soon UR.io users will have the option of a PrivateInference-bundled subscription service too. Say it with us: Private is Better
7
14
2
48
2,668
Evgenij2050 retweeted
Many people’s most honest conversations this year were with chatbots owned by companies that have their billing address. @P_Inference keeps the them separate: wallet payment, no account, and prompts encrypted the prompt before they're sent. V6 launch partner, now in open beta.
🔓 Private Inference is now in Open Beta No more whitelists - login now at privateinference.gg 🧠 Frontier and open models side by side, in one chat and one OpenAI-compatible API. Fund a session once from your wallet. Each request is held at its maximum cost, and you pay only for what you actually use. 💸 Pay your way: · USDC on Base · cUSDC on Aztec, privately · ZEC from any wallet No model accounts, no cards, no monthly plans. 🔐 You decide how private each request is: · Zero data retention, the default · Standard routes, usually cheaper · Private execution (beta): your browser encrypts the prompt to a model in an attested TEE and checks the attestation itself 🤝 Bring your friends. Your referral code sits at the top of every page, and friends you bring in earn you balance credit. 🚀 We ship every week: privateinference.gg/changelo…
6
11
59
5,125
Your coding agent can now run more privately: Private Inference added the Responses API, the protocol Codex and other agents speak. We don't store prompts or replies, and ZDR sends each request only to suppliers that keep nothing. What that means for coders: one provider block in your Codex config, then codex --profile pi. Your agent keeps its tools and workflow, and you pick Standard or ZDR per request. docs.privateinference.gg/api…
4
9
27
708
🕶️ Pay for AI without revealing who you are. Prompt it without revealing what you asked.
4
13
3
36
783
What’s new since last week 🚀 (and something big on Friday 👀) 📎 Attach files to a message: notes, code, CSVs, logs (up to 10 files, 512 KB each). Drop them on the prompt or use the paperclip, and the model reads them with your message. 🔎 Web search in chat: tap the magnifier and the model gets up to ten pages from the web, cited right in the reply. Hover a citation to preview the page. Works with every model at the standard tier. 📚 Docs are live at docs.privateinference.gg: who sees what at each privacy level, how private execution is checked, what we keep, payments and the API. 🚪 A new Start page: “Open app” takes you straight to sign-in, “Get an API key” drops you into your keys. 💬 Questions or feedback? There’s now a Discord link in the header. 🛡️ cUSDC on Aztec: if your cUSDC is sitting on your public balance, we now tell you that and how to move it to private, instead of “Balance too low.” 🧮 Math that looks like math: nested fractions and subscripts like d₁ now render properly. And the big one 👇 Private Inference leaves closed beta and opens to everyone (Open Beta) on Friday 23 October. No invite needed. One balance, every model, in chat or by API. Top up in USDC on Base, private cUSDC on Aztec or ZEC, and pay only for what you use. Changelog: privateinference.gg/changelo…
3
14
3
32
1,985
Private AI. One API. Private stablecoin payments. Access leading models with cUSDC on Aztec. Your next prompt deserves more privacy. 🛡️ Join us → privateinference.gg/?ref=MyN… Let's go !!!👍👍👍
🤖 Made with AI
2
66
Private AI. Private money. cUSDC on the @aztecnetwork → privateinference.gg No KYC, no logs — just inference. Join the private AI stack at an early stage. privateinference.gg/?ref=MyN…
🤖 Made with AI
4
106
AI that doesn’t know who you are. Private inference powered by @aztecnetwork. Pay with cUSDC. Access GPT & Claude. No KYC. No logs. The future is private → privateinference.gg/?ref=MyN… Let's go!!!
🤖 Made with AI
1
9
376
Evgenij2050 retweeted
Replying to @GalacticaNet
@GalacticaNet just launched a direct competitor to Venice AI, but with one major advantage: private payments. when you use ChatGPT, Claude, or other AI services, you might be revealing much more about yourself than you realize. your health, finances, work, personal life. all of this information can be processed and stored on corporate servers, with your conversations linked to your account. these companies get access to some of your most personal information, and if a data breach happens, your conversations could end up in the wrong hands. and now that ads have arrived in ChatGPT, the question of how much personal data we share with AI companies is becoming even more relevant. why share all of this information when you can use the same powerful AI models much more privately? that's exactly what @P_Inference is building. → all your favorite AI models in one place. GPT, Claude, Gemini, and more. no need to create separate accounts or pay for multiple subscriptions. → pay only for what you actually use. fund your balance with crypto through x402 and pay based on your actual AI usage. no monthly subscriptions, no credit cards, no paying for limits you never use. → keep your conversations private. in Private mode, your prompts are encrypted directly in your browser before being processed inside a Trusted Execution Environment (TEE). the goal is to keep even Private Inference from reading your conversations. sounds familiar? that's because Venice AI offers pretty much the same features. but there's one major problem Venice hasn't solved yet. even if your AI conversations are encrypted, paying for those services on a public blockchain still leaves a trail. your wallet address, payment amounts, and transaction history are publicly visible. this doesn't reveal your actual prompts, but it allows others to track your spending and potentially link your wallet to your AI usage. Private Inference solves this with @aztecnetwork. thanks to private payments on Aztec, outside observers can't see the sender, recipient, or amount of a private transfer onchain. so you get: → private AI conversations → encrypted requests → access to multiple AI models → private crypto payments all in one platform. Private Inference is currently in closed beta, with access available through invite codes. the private AI era is just getting started 🛡️
If ethereum:0xa27ec0006e59f245217ff08cd52a7e8b169e62d2 has caught your attention, take a look at what Galactica is building on it. @P_Inference lets you pay for AI privately through Aztec and your questions all the way to the model. Talk to AI in confidence. Sign up for access: privateinference.gg/aztecwhi…
6
6
1
38
1,191
Evgenij2050 retweeted
If ethereum:0xa27ec0006e59f245217ff08cd52a7e8b169e62d2 has caught your attention, take a look at what Galactica is building on it. @P_Inference lets you pay for AI privately through Aztec and your questions all the way to the model. Talk to AI in confidence. Sign up for access: privateinference.gg/aztecwhi…
9
8
3
98
5,395
We’re proud to be using @humntech bridged USDC (cUSDC) to power private payments on Private Inference. 🛡️ Most AI services want your card, your email and your payment history. With cUSDC on @aztecnetwork private rail, you can fund your account without exposing any of that on chain. What this means for you: → 💰 Top up with cUSDC. The sender, recipient and amount stay off chain. → 🚫 No card, no monthly plan, no separate accounts with each model supplier. → 🔑 One balance and one API key give you access to GPT and Claude models. → 📊 Each request is held at its maximum cost, then charged only for actual usage. Unused funds are released right away. A big thank you to the @humntech team. Their work on bridged USDC is what makes private stablecoin payments for AI possible.
5
5
3
51
2,808
Zero-knowledge proofs meet large language models. Inference that verifies without revealing. Payments that settle without exposing. This is privateinference.gg privateinference.gg/?ref=MyN…
🤖 Made with AI
5
151
What we’ve done since Oct 6 (closed-beta launch) 🚀 It’s been a busy week! Here’s what changed since the doors opened: 👀 Replies you can actually see: when a reply contains HTML or SVG, you now get the finished thing (a widget, a card, a chart) right in the chat, with the code one click away. It runs in a locked-down frame that can’t touch your account or chats. ⚡ No more freezing: long chats stay smooth while a reply streams in, and bold, code and lists format as they arrive instead of showing up as stray symbols. 🔑 API in under a minute: a new key now comes with a ready-to-run command (macOS, Linux or Windows) and a button that fires a test request. 💳 Fewer payment and sign-in headaches: cancelling a payment no longer throws an error, Aztec payments work again, and when sign-in fails you’re told what to do next instead of getting an error code. 🤝 Invite your friends: as of today, everyone on the whitelist can invite up to 10 people, and everyone who gets invited can invite up to 5 more. Your referral code and a copy-link button live at the top of every page. 🎨 Code that’s easier on the eyes: syntax highlighting in both dark and light themes. ✨ And a lot more: mobile menu, smarter top-ups, a handful of smaller fixes and polish all over the place. Thanks to everyone testing and sending feedback, it’s shaping what we build next 💙 Stay posted in the Changelog: privateinference.gg/changelo…
4
10
2
42
1,256
Building private inference for people and agents GPT, Claude & more through one API key Pay privately with stablecoins via @P_Inference @aztecnetwork using x402 privateinference.gg/?ref=MyN…
🤖 Made with AI
2
61
Have you had a chance to try out privateinference.gg yet?
privateinference.gg is now live for private whitelisted beta phase 1. What it is, why it exists, and how the whitelist works. A thread 🧵
6
6
1
24
767