OmniVoice-Triton-API

OpenAI-compatible text-to-speech (TTS) API that accelerates the open-source OmniVoice model with Triton kernel fusion and CUDA Graph optimization, delivering…

  • AI
  • Productivity
  • Tools & Utilities
OpenAI-compatible text-to-speech (TTS) API that accelerates the open-source OmniVoice model with Triton kernel fusion and CUDA Graph optimization, delivering 646-language speech synthesis, voice cloning, and voice design at up to 3.4× the speed of stock OmniVoice (~170 ms per utterance) for NVIDIA GPU users.

Platform availability

Available on 1 of 10 platforms

  • Unraid
  • TrueNAS (not listed)
  • Umbrel (not listed)
  • ZimaOS (not listed)
  • Proxmox (not listed)
  • Helm (not listed)
  • Mac (not listed)
  • Windows (not listed)
  • Linux (not listed)
  • Any server (not listed)

Health score?A 0-100 score computed nightly from four factors: maintained (40%: recent commits and steady activity), popular (25%: GitHub stars, log scale), easy to install (20%: platforms, install notes, setup guides, Docker image) and light to run (15%: minimum RAM, ARM support). When the source code has been checked, a fifth factor, security check (20%: automated scan for committed secrets, unsafe desktop-app settings, known-vulnerable dependencies and project hygiene; any serious finding caps it at 20), is added. Factors without data are left out and the rest are rescaled.

35/100 Low

Maintained
59last commit 164 days ago
Popular
102 GitHub stars
Easy to install
351 install platform, install notes, Docker image
Light to run
108 GB minimum RAM

Availability: platform listings confirmed today, links checked 19 days ago

GitHub stars
2
Open issues
0
Last commit
2026-06-10
Activity
stale

Checked 6 days ago - source: GitHub (hsiang-han/OmniVoice-Triton-API)

Resources & Compatibility

Minimum requirements: 8 GB RAM, 4 CPU core(s), 3 GB disk.

Checked 54 days ago - source: Docker Hub image tags and published docs

First-install notes

  • Install the package before running it Run pip install omnivoice, or clone the repo and run uv sync, to get the basic app commands. Without that step, commands like omnivoice-demo will not exist on your machine. source
  • Open port 8001 for the demo page Start the demo with omnivoice-demo --ip 0.0.0.0 --port 8001; port 8001 is the numbered gate your browser uses, and 0.0.0.0 means it can be reached from other devices. If you do not keep port 8001 accessible, the web page may not load from another machine. source
  • Install the matching PyTorch build If you have an Apple M-series Mac, run pip install torch==2.8.0 torchaudio==2.8.0, or use the Intel Arc XPU command for Intel Arc GPUs. The wrong PyTorch (the math engine that runs the model) build can make the model run slowly or fail to use the GPU correctly. source
  • GPU and first-start setup Requires an NVIDIA GPU with driver version 550 or later (~2 GB VRAM). On first start the ~2 GB model downloads automatically from Hugging Face. China users: set HF_ENDPOINT to https://hf-mirror.com for faster downloads. source

Fetched 47 days ago - each note links to its source

Alternatives

No alternatives collected yet.

Common questions

Can I point existing OpenAI TTS clients at this service?

Yes, the project exposes an OpenAI-compatible TTS API.

What step is required before omnivoice-demo commands exist?

Install the package first with pip install omnivoice, or clone the repo and run uv sync. Without that step, commands like omnivoice-demo will not exist on your machine.

How do I make the OmniVoice demo page reachable from another machine?

Start the demo with omnivoice-demo --ip 0.0.0.0 --port 8001 and keep port 8001 accessible. If port 8001 is not accessible, the web page may not load from another machine.

Which PyTorch build should I install for Apple M-series Macs or Intel Arc GPUs?

On an Apple M-series Mac, run pip install torch==2.8.0 torchaudio==2.8.0. For Intel Arc GPUs, use the Intel Arc XPU command, because the wrong PyTorch build can make the model run slowly or fail to use the GPU correctly.

Does the Triton acceleration require extra VRAM?

No, the Triton kernels fuse multiple HBM round-trips into a single pass and reduce memory traffic without any additional VRAM usage.

Answers sourced from github.com, github.com, github.com

Community

Collected 49 days ago - public community threads