Lutheos · Private inference API

Lutheos

Lutheos is a 'Private AI Inference' API; larprouter.com now redirects to lutheos.com. It serves open models self-hosted by its GPU partner with zero data retention, either abliterated or as released. One endpoint speaks both the OpenAI and the Anthropic API. The models listed are DeepSeek V4.1 Flash 1017 Abliterated, DeepSeek V4.1 Flash 1017 and GLM 5.3 Flash. Pricing is pay-as-you-go from a prepaid balance with no monthly fee, and crypto payment is accepted. DeepSeek V4.1 Flash 1017 Abliterated costs $0.08 input and $0.32 output per 1M tokens, with a 1M context and 128K max output. GLM 5.3 Flash costs $0.08 input and $0.42 output per 1M tokens.