@engyai

engy.ai provides OpenAI-compatible inference API for frontier open-source LLMs (DeepSeek, GLM, Kimi, Qwen) with cryptographic proof of correct inference.

They claim competitive pricing (30-68% cheaper than alternatives) and offer zero data retention with agentic-native design.

Recently launched models include DeepSeek-V4.1-Flash and GLM-5.3, with mining infrastructure for select models.

Small following (<3k) with 2-tier scout convergence suggests early discovery, though verification claims and economic model need validation.

Green flags: Real technical differentiation: cryptographic proof-of-inference is a legitimate problem in decentralized AI · Working product on mainnet serving multiple frontier models (DeepSeek, GLM, Kimi, Qwen) · Competitive pricing advantage documented across multiple models vs established providers · Two Tier A scouts converging (0xKant_ 66/100, ZenRacc00n 78/100) · Small follower count (2,658) with substance: active API, Discord community, frequent model updates

Red flags: Account created Oct 2020 but appears dormant until recent activity (raises questions about project timeline) · No token mentioned despite mining references (unclear tokenomics/incentive structure) · Limited technical documentation on verification mechanism in provided data · Heavy reliance on third-party model access (Kimi, DeepSeek) creates dependency risk