OpenAI-compatible serving on your GPUs, fronted by QUIC.
/v1/chat/completions and /v1/audio/* — point the client you already use at the edge and go.
A QUIC front-proxy at the edge roughly halves WAN time-to-first-token versus HTTP/1 — measured, not marketed.
Models deploy per tenant on Kubernetes through the control plane, with metering and routing handled by the platform.
Self-hosted models plug straight into agents — including Qwen Omni speech-to-speech replacing an entire ASR/LLM/TTS pipeline.
Telequick runs every layer of voice — you bring the models.