# Ksuriuri/index-tts-vllm

Added vLLM support to IndexTTS for faster inference.

Repository: https://github.com/Ksuriuri/index-tts-vllm
Canonical: https://ross.abutalabs.com/products/index-tts-vllm
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2026-04-13T10:31:22+00:00

## Health v2 (maintenance only)
Score: 58/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 77, release rhythm 47, longevity 34
- inputs: {"age_days": 487, "days_push": 142, "days_rel": 142, "gap_med": 211, "n_releases_24m": 2}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1235, forks 174 (observed 2026-08-28T04:04:04.915441+00:00)

## What it is
A reimplementation of IndexTTS's GPT model inference using vLLM, providing significantly faster text-to-speech generation with a web UI and FastAPI-based API server. It supports Index-TTS v1, v1.5, and v2 model weights with OpenAI-compatible API options.

## Use cases
- run index-tts with faster inference using vllm
- self-host a text-to-speech api server
- serve tts with high concurrency on a single gpu
- deploy indextts2 with a webui
- speed up gpt-based tts decoding
- expose openai-compatible speech synthesis endpoint

## When to choose
- you want much lower RTF and higher concurrency for IndexTTS on an NVIDIA GPU
- you need a self-hosted TTS service with REST or OpenAI-compatible APIs
- you want a ready webui and prepackaged model weights for easy deployment

## When to avoid
- you don't have a CUDA-capable NVIDIA GPU
- you need non-IndexTTS TTS models
- you need fully parallel non-GPT pipeline stages - s2mel inference is still serial and slow

## Facets
- artifact type: application
- maturity: active
- function: tts, llm-inference, http-server, gpu-computing
- domain: speech-processing, artificial-intelligence, large-language-models
- platform: python
- tags: vllm, text-to-speech, indextts, inference-acceleration, fastapi, voice-cloning, webui, gpu, linux, docker

## Member repositories
- Ksuriuri/index-tts-vllm (main) score 58

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:04.915441+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T08:22:02.659568+00:00, confidence not recorded.
  - readme: https://github.com/Ksuriuri/index-tts-vllm (fetched 2026-08-28T04:04:04.915441+00:00, sha 4a4ac8ce8cbb)
- Data as of 2026-08-30T08:39:29.467469+00:00.