Test my new LLM inference engine and get 100M free tokens and then 60% cheaper inference

Open-weight inference with clear per-token pricing. Start with Qwen 3.8 27B, request custom models, or deploy entirely on AWS or GCP.