Vitalik Buterin Tests Private AI Setup for Personal Health Recommendations

Ethereum co-founder Vitalik Buterin has been experimenting with a private AI setup to generate personalized diet and exercise recommendations while limiting the amount of personal information sent to remote models.

The setup uses Alibaba's Qwen3.8-Flash-Next as the local model, which handles the first layer of privacy by constructing queries itself instead of sending original wording and full personal context to remote AI systems.

The second layer uses zkAPI, a system developed by the Open Anonymity Project in collaboration with the Ethereum Foundation, to separate payments from individual AI requests.

The third layer uses Tor to hide the user's normal IP address from the services receiving network requests.

Buterin said Tor latency remains 10-100 times higher than desirable, making request-by-request unlinking inefficient in current tests.

He also noted that the local model is currently running at approximately 20-30 tokens per second, but he believes local inference would only begin to feel fast at more than 100 tokens per second.