---
title: "Step-Audio-R1 by StepFun AI | AIDeveloper44"
url: https://aideveloper44.com/product/step-audio-r1-696b3f209dfae48dc5aa7cc8
description: "Step-Audio-R1 is an advanced audio language model developed by StepFun AI, designed to enhance audio reasoning capabilities by grounding its reasoning in acoustic features. It introduces Modality-Grounded Reasoning Distillation (MGRD), an iterative training framework that shifts the model's reasoning from textual abstractions to acoustic properties, effectively addressing the 'inverted scaling' problem where performance degrades with longer reasoning. This model has demonstrated superior performance across various audio understanding and reasoning benchmarks, surpassing models like Gemini 2.5 Pro and achieving results comparable to Gemini 3 Pro."
lang: en
---

← all releases (https://aideveloper44.com/releases)

StepFun AI · Nov 29, 2025 · v1.0 · open source

# Step-Audio-R1

Step-Audio-R1 is an advanced audio language model developed by StepFun AI, designed to enhance audio reasoning capabilities by grounding its reasoning in acoustic features. It introduces Modality-Grounded Reasoning Distillation (MGRD), an iterative training framework that shifts the model's reasoning from textual abstractions to acoustic properties, effectively addressing the 'inverted scaling' problem where performance degrades with longer reasoning. This model has demonstrated superior performance across various audio understanding and reasoning benchmarks, surpassing models like Gemini 2.5 Pro and achieving results comparable to Gemini 3 Pro.

LLMs ML Testing

visit website: https://huggingface.co/stepfun-ai/Step-Audio-R1

overview

key features

- 01 Chain-of-Thought (CoT) Reasoning: Unlocks CoT reasoning capabilities in audio language models, generating reasoning chains grounded in acoustic features.
- 02 Modality-Grounded Reasoning Distillation (MGRD): An iterative training framework that shifts reasoning from textual abstractions to acoustic properties.
- 03 Superior Performance: Surpasses Gemini 2.5 Pro and is comparable to Gemini 3 across major audio reasoning tasks.
- 04 Versatility: Covers speech, environmental sounds, and music domains.
- 05 Open Source: Released under the Apache 2.0 license, promoting accessibility and collaboration.

related products browse all → (https://aideveloper44.com/releases)

SINQ open source SINQ (Sinkhorn-Normalized Quantization) is a novel, fast, and high-quality quantization method designed to make any Large Language Model (LLM) smaller while preserving accuracy. It offers a plug-and-play, model-agnostic technique that delivers state-of-the-art performance for LLMs without sacrificing accuracy.: https://aideveloper44.com/product/sinq-698c2e4a4e5c16e85ff33ab5
DetectFlow open source DetectFlow is an open-source cybersecurity detection platform developed by SOC Prime. It leverages artificial intelligence to enhance the detection of cyber threats in real-time, enabling security operations teams to identify and respond to attacks more effectively.: https://aideveloper44.com/product/detectflow-699d62f9abbca3d71a8a85d0
Protenix open source Protenix is an open-source, trainable PyTorch implementation of AlphaFold 3, designed for high-accuracy biomolecular structure prediction. It aims to advance accessible and extensible research tools for the computational biology community. ([github.com](https://github.com/bytedance/Protenix?utm_source=openai)): https://aideveloper44.com/product/protenix-698868b72dc0c583b95fa31a

## Structured data

```json
[
  {
    "@context": "https://schema.org",
    "@type": [
      "WebSite",
      "NewsArticle"
    ],
    "name": "AIDeveloper44",
    "url": "https://aideveloper44.com",
    "isAccessibleForFree": true,
    "isPartOf": {
      "productID": "CAowsfXgCw:openaccess",
      "@type": [
        "Product"
      ]
    }
  },
  {
    "@context": "https://schema.org",
    "@type": "Organization",
    "name": "AIDeveloper44",
    "url": "https://aideveloper44.com",
    "logo": "https://qtrypzzcjebvfcihiynt.supabase.co/storage/v1/object/public/base44-prod/public/696b2bd107af20deac4cc200/9f77c0b91_a-digital-illustration-of-a-white-llama-_j4dyH5xoRnKf_3rtw-mvEQ_OdqL2BrSTtKHjJAuhTp_zg.png"
  },
  {
    "@context": "https://schema.org",
    "@type": "BreadcrumbList",
    "itemListElement": [
      {
        "@type": "ListItem",
        "position": 1,
        "name": "Home",
        "item": "https://aideveloper44.com/"
      },
      {
        "@type": "ListItem",
        "position": 2,
        "name": "product",
        "item": "https://aideveloper44.com/product"
      },
      {
        "@type": "ListItem",
        "position": 3,
        "name": "Product Detail",
        "item": "https://aideveloper44.com/product/step-audio-r1-696b3f209dfae48dc5aa7cc8"
      }
    ]
  },
  {
    "@context": "https://schema.org",
    "@type": "SoftwareApplication",
    "name": "Step-Audio-R1",
    "description": "Step-Audio-R1 is an advanced audio language model developed by StepFun AI, designed to enhance audio reasoning capabilities by grounding its reasoning in acoustic features. It introduces Modality-Grounded Reasoning Distillation (MGRD), an iterative training framework that shifts the model's reasoning from textual abstractions to acoustic properties, effectively addressing the 'inverted scaling' problem where performance degrades with longer reasoning. This model has demonstrated superior performance across various audio understanding and reasoning benchmarks, surpassing models like Gemini 2.5 Pro and achieving results comparable to Gemini 3 Pro.",
    "provider": {
      "@type": "Organization",
      "name": "StepFun AI"
    },
    "offers": {
      "@type": "Offer",
      "price": "Free"
    }
  },
  {
    "@context": "https://schema.org",
    "@type": "Organization",
    "name": "AINews.sh",
    "alternateName": "AI News",
    "url": "https://ai-dev-pulse-ac4cc200.base44.app",
    "logo": "https://ai-dev-pulse-ac4cc200.base44.app/logo.png",
    "description": "Developer-focused AI product releases, reviews, and news. Your source for the latest AI tools, open-source models, and platform updates.",
    "foundingDate": "2026",
    "slogan": "Stay ahead with AI product intelligence",
    "sameAs": [],
    "knowsAbout": [
      "Artificial Intelligence",
      "Machine Learning",
      "Large Language Models",
      "AI Tools",
      "Open Source AI",
      "AI Frameworks",
      "Developer Tools"
    ]
  }
]
```