# Auritus — just-in-time narration for the web > Open-source embed that turns page content into spoken audio with a local GPU worker and cloud fallback. ## region-001 Open source narration infrastructure ## region-002 Let every page speak. ## region-003 Auritus turns the readable text on any page into on-demand audio — with a small embed, open-model workers you run, and an AWS safety net behind them. ## region-004 Embed ## region-005 script tag on a publisher page ## region-006 Local + cloud ## region-007 compute paths: local first, cloud fallback ## region-008 Open models ## region-009 models and deployment choices ## region-010 Hear it for yourself ## region-011 What Auritus does ## region-012 A 30-second introduction, narrated by Auritus ## region-013 Press play. What you hear is this page reading itself. ## region-014 Auritus turns any article into audio on demand, from a single script tag — no pre-recording, no studio, no per-article cost. ## region-015 You choose the voice model. You decide where the text goes. Your own GPU does the work, and AWS picks up the slack when it cannot. ## region-016 It's open source, and every article on your site can speak, with the choices that matter still yours. ## region-017 Narrated by Auritus with Kokoro (af\_heart) — the same embed documented in Usage. ## region-018 What just happened ## region-019 One script tag, and the page reads itself. ## region-020 Control is the product ## region-021 A narration layer that adapts to your stack. ## region-022 Build a listening experience without treating your publisher content and production choices as somebody else's defaults. ## region-023 Your data path ## region-024 Narration on infrastructure you choose. ## region-025 Keep page text in a stack you operate. Audio stays private by default and is delivered through short-lived access. ## region-026 Your model ## region-027 Use an open model. Tune the tradeoff. ## region-028 Choose the voice backend and parameters that fit your quality, latency, and cost targets instead of accepting one fixed service. ## region-029 Your compute ## region-030 Start local. Fall back when it matters. ## region-031 A GPU worker you run can claim ordinary work first. AWS Batch is there when that worker is unavailable or the queue needs help. ## region-032 Designed for a reliable handoff ## region-033 Your GPU when it is there. AWS when it is not. ## region-034 A local worker races to claim the job. If nobody claims it during the configured window, a workflow submits an AWS Batch GPU job. The conditional claim means only one worker finishes the audio. ## region-035 Content-hash deduplication keeps repeated requests efficient. ## region-036 Private audio is returned with short-lived access. ## region-037 Per-site quotas, budgets, and a kill switch limit cloud work. ## region-038 A short path to first audio ## region-039 Deploy, register, embed. ## region-040 You can run the whole stack yourself. If you want help, we can operate a deployment in your account or build something for you with Auritus. ## region-041 You choose how much we help ## region-042 MIT licensed. Deploy it yourself. Anthus does nothing and charges nothing. ## region-043 Deployed in your AWS account. We run the setup session with you, then operate it and keep it updated. We update our own deployments first. ## region-044 Hosted narration is not offered yet; narration runs on GPU you control. ## region-045 We build something for you with Auritus and operate it if you want. A fixed-scope pilot first, then monthly. ## region-046 Straight answers ## region-047 What happens if I stop paying? ## region-048 Dedicated deployments stay in your account untouched; we stop operating them and delete nothing. ## region-049 Who pays for what? ## region-050 Dedicated: you pay AWS, including GPU, and any model providers directly; Anthus charges only the operations fee. Engagement: stated in the scope. ## region-051 What does dedicated actually mean? ## region-052 Your data never leaves your account. Your own model contracts and keys. Your compliance boundary. No noisy neighbors. You keep running if Anthus disappears. The monthly fee buys updates, monitoring, and a team that has run the same deployment first: we update our own deployments before yours. ## region-053 Can I start with open source and move to Dedicated? ## region-054 Yes, and the reverse. ## region-055 Security architecture ## region-056 No IAM keys on the worker. None to leak. ## region-057 The usual way to let a machine call your service is a long-lived IAM access key, scoped wider than anyone intended and rotated on a schedule someone has to remember. Auritus issues none. An operator signs in once — with a Cognito account, or your own identity provider — and the worker then carries a one-hour token that renews itself and cannot call AWS APIs at all. The risk is removed by the architecture rather than managed by procedure. ## region-058 The security page states plainly which parts are enforced today and which are still being built, including the remaining JWT validation work. ## region-059 Open-source, just-in-time narration for the web.