Chat & RAG
chat (multimodal, tools, vision), code (dedicated coder), and local — a sovereign CPU chat model
whose prompts never leave our box.
ollalink API
One OpenAI-compatible API for chat and real-time speech — TTS, transcription, and live translation.
What is ollalink?
Section titled “What is ollalink?”ollalink is a partner API for conversational AI and real-time speech. Everything speaks the OpenAI-compatible REST + SSE surface, so most OpenAI SDKs work by changing only the base URL and key.
Speech in
Transcribe audio with whisper (CPU) or gpu-asrhq (whisper large-v3 on GPU) — ~99 languages, timestamps.
Speech out
Natural TTS — Soniox voices plus OmniVoice expressive voices — with engineer-style verbalization of IPs & codes.
Live translate
Stream mic audio over WebSocket and get partial/final transcripts + translated captions + spoken audio.
Base URL
Section titled “Base URL”https://ai.ollalink.com/v1Send your partner key as a Bearer token on every request. See Getting started.