# stepfun-ai/Step-Audio2

Step-Audio 2 is an end-to-end multi-modal large language model designed for industry-strength audio understanding and speech conversation.

Repository: https://github.com/stepfun-ai/Step-Audio2
Canonical: https://ross.abutalabs.com/products/step-audio2
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2026-03-16T04:06:21+00:00

## Health v2 (maintenance only)
Score: 50/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 72, release rhythm 35, longevity 29
- inputs: {"age_days": 414, "days_push": 170, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1503, forks 113 (observed 2026-08-28T04:04:54.635650+00:00)

## What it is
Step-Audio 2 is an end-to-end multimodal large language model for industry-strength audio understanding and speech conversation, with open-weight mini variants and inference examples. It supports ASR, TTS, paralinguistic understanding, and tool calling from audio input.

## Use cases
- build a voice assistant that understands speech end-to-end
- transcribe speech with paralinguistic awareness
- generate natural speech responses from an LLM
- run audio chat models with vLLM
- evaluate audio models on tool-calling benchmarks

## When to choose
- you need an open-weight audio-language model for speech conversation
- you want combined audio understanding and generation in one model
- you need Apache-2.0 licensed speech AI

## When to avoid
- you need a lightweight CPU-only speech solution
- you only need simple ASR without LLM conversation
- you lack GPU resources for large model inference

## Facets
- artifact type: library
- maturity: active
- function: speech-recognition, tts, llm-inference, machine-learning
- domain: speech-processing, large-language-models, artificial-intelligence
- platform: python
- tags: audio-understanding, speech-conversation, multimodal-llm, voice-cloning, asr, vllm, audio, gpu, linux

## Member repositories
- stepfun-ai/Step-Audio2 (main) score 50

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:54.635650+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 2026-08-30T04:32:50.502742+00:00, confidence not recorded.
  - readme: https://github.com/stepfun-ai/Step-Audio2 (fetched 2026-08-28T04:04:54.635650+00:00, sha 4132302d7f2b)
  - registry_pypi: https://pypi.org/pypi/step-audio2/json (fetched 2026-08-29T11:37:36.626501+00:00, sha 29ecfa6089f7)
- Data as of 2026-08-30T08:39:29.467469+00:00.