They do verifiable inference on TEEs for the open source models. The anthropic ones I think they basically proxy for you (also via trusted TEE) so that it cant be tied to you. VPN for LLM inference so to speak.

Can you link to the verifiable inference method?