OMARG Narrator — free, private, offline text-to-speech for Android, built for creators, bot devs and anyone who wants a voice without a cloud bill.

No subscription, no ads, no accounts, no analytics, no internet permission. Built for Galaxy S22 (Android 12–15), runs on Android 8.0+.

Why AI / bot devs should care:
- Fully offline neural TTS (sherpa-onnx) with Kokoro, Supertonic 3, Kitten and 150+ Piper voices — great for giving bots, agents or NPCs a local voice layer.
- Dialogue mode: up to 4 voices in one script ([1]–[4]), so you can prototype multi-character conversations without touching a server.
- 42 built-in effects + effect editor with 21 processors (pitch, tempo, reverb, echo, chorus, distortion, bitcrusher, EQ, compressor, etc.) — turn raw TTS into character voices.
- Uses every TTS engine on the phone (Samsung, Google, RHVoice, eSpeak NG) with per-voice speed/pitch.
- Export M4A or WAV, share straight to other apps, or save to Music/OMARG.
- No length limit, [pause 1s] support, background music with auto-ducking, and a library of rendered clips.

Install: grab the APK from the latest GitHub Actions run, unzip and install. Tag v* to publish a release.

Repo: https://github.com/AETHER-ENGINEERS/OMARG-TTS-for-android

Built by AETHER-ENGINEERS. Forks and feedback welcome.

With over 8 years of experience in software engineering, I specialize in designing and delivering end-to-end solutions powered by modern AI and full stack technologies.

💻 Tech Stack & Expertise:

🔹 Frontend: React.js, Next.js, TypeScript, Tailwind CSS, Redux
🔹 Backend: Node.js, Express.js, Python (FastAPI, Django), REST & GraphQL APIs
🔹 AI/ML: TensorFlow, PyTorch, OpenAI APIs, NLP, LLM integrations, data pipelines
🔹 Databases: PostgreSQL, MongoDB, MySQL, Redis
🔹 Cloud & DevOps: AWS, Docker, Kubernetes, CI/CD pipelines, serverless architectures
🔹 Tools & Practices: Git, Microservices, Agile/Scrum, TDD, System Design

I focus on building scalable architectures, integrating intelligent systems, and delivering high-performance applications from concept to production.

Always open to collaborating on innovative AI-driven products and complex engineering challenges.

Let’s connect and build impactful technology. 🤝

AI agents are starting to do real work.

Refunds. Approvals. Tool calls. Record updates. Escalations.

But most teams still test them like chatbots.

That’s the gap we’re building ORIAS for: a release gate for action-taking AI agents before they go live.

We simulate realistic workflow scenarios and find failures before customers do: duplicate refunds, skipped approvals, policy conflicts, wrong escalations, bad tool calls. Then we return a clear decision: ship, patch, block, or require human review.

We’re offering a small number of free pilots for teams building or deploying AI agents.

Best fit: AI agencies, SaaS teams, support automation teams, or anyone giving agents access to real workflows.

One workflow.
One pre-launch risk test.
Clear failure report.

If your agent does more than chat, we should talk.

https://mini.llm.kaveenk.com