Close Menu
  • Home
  • Opinion
  • Region
    • Africa
    • Asia
    • Europe
    • Middle East
    • North America
    • Oceania
    • South America
  • AI & Machine Learning
  • Robotics & Automation
  • Space & Deep Tech
  • Web3 & Digital Economies
  • Climate & Sustainability Tech
  • Biotech & Future Health
  • Mobility & Smart Cities
  • Global Tech Pulse
  • Cybersecurity & Digital Rights
  • Future of Work & Education
  • Trend Radar & Startup Watch
  • Creator Economy & Culture
What's Hot

His Excellency Sheikh Nahyan bin Mubarak Al Nahyan crowns winner of Longines Grand Prix class at FBMA Worldwide Present Leaping Cup 2026

January 18, 2026

Mario’s Tiny Odyssey Has Him Conquering the Apple Watch

January 18, 2026

Beijing Completes Rocket Road, China’s First Shared Testing and R&D Manufacturing Base for Industrial Spaceflight

January 18, 2026
Facebook X (Twitter) Instagram LinkedIn RSS
NextTech NewsNextTech News
Facebook X (Twitter) Instagram LinkedIn RSS
  • Home
  • Africa
  • Asia
  • Europe
  • Middle East
  • North America
  • Oceania
  • South America
  • Opinion
Trending
  • His Excellency Sheikh Nahyan bin Mubarak Al Nahyan crowns winner of Longines Grand Prix class at FBMA Worldwide Present Leaping Cup 2026
  • Mario’s Tiny Odyssey Has Him Conquering the Apple Watch
  • Beijing Completes Rocket Road, China’s First Shared Testing and R&D Manufacturing Base for Industrial Spaceflight
  • One Maker Turned 3,200 Matches right into a Single Working Large
  • NVIDIA Releases PersonaPlex-7B-v1: A Actual-Time Speech-to-Speech Mannequin Designed for Pure and Full-Duplex Conversations
  • Fizz Cellular launches free eSIM trial
  • AMD says it can attempt to hold GPU costs down amid RAM scarcity
  • StepFun’s Open-Supply Speech Mannequin Step-Audio-R1.1 Tops International Benchmark Rankings
Sunday, January 18
NextTech NewsNextTech News
Home - Asia - StepFun’s Open-Supply Speech Mannequin Step-Audio-R1.1 Tops International Benchmark Rankings
Asia

StepFun’s Open-Supply Speech Mannequin Step-Audio-R1.1 Tops International Benchmark Rankings

NextTechBy NextTechJanuary 18, 2026No Comments2 Mins Read
Share Facebook Twitter Pinterest LinkedIn Tumblr Telegram Email Copy Link
Follow Us
Google News Flipboard
StepFun’s Open-Supply Speech Mannequin Step-Audio-R1.1 Tops International Benchmark Rankings
Share
Facebook Twitter LinkedIn Pinterest Email



In keeping with an announcement from StepFun on January 15, its open-source mannequin Step-Audio-R1.1 has ranked No.1 worldwide on a significant worldwide analysis leaderboard.

IMG_4778.jpeg

StepFun mentioned that the Synthetic Evaluation Speech Reasoning benchmark—one of the vital authoritative third-party evaluations for native audio fashions—has been up to date, with Step-Audio-R1.1 taking the highest place. The benchmark is extensively thought to be a key trade commonplace for assessing fashions that course of uncooked audio immediately and carry out advanced logical reasoning, with core analysis metrics together with accuracy and first-token latency.

In keeping with StepFun, similar to giant language fashions, speech fashions should possess robust reasoning capabilities to allow higher-level intelligence and extra pure human–machine interplay. With an accuracy rating of 96.4%, Step-Audio-R1.1 outperformed main fashions akin to Grok, Gemini, and GPT-Realtime, setting a brand new all-time document on the benchmark. By way of the general steadiness between efficiency and velocity, Step-Audio-R1.1 considerably outperforms comparable speech fashions.

IMG_4779.jpeg

Step-Audio-R1, initially launched by StepFun, is acknowledged because the world’s first open-source native speech reasoning mannequin. It’s able to end-to-end speech understanding and reasoning with out introducing further latency, enabling the system to “assume as quickly because it hears,” very like a human listener.

Its core capabilities embrace:

  • Superior speech reasoning
  • Actual-time response efficiency
  • Scalable chain-of-thought (CoT) reasoning within the audio area

Step-Audio-R1.1, the most recent upgraded model, additional enhances real-time conversational efficiency and complicated speech reasoning capabilities. A full real-time speech API is scheduled for launch in February. The presently accessible chat mode already integrates the R1.1 core, supporting streaming inference that permits the mannequin to assume and communicate concurrently.

Supply: IT Residence

Elevate your perspective with NextTech Information, the place innovation meets perception.
Uncover the most recent breakthroughs, get unique updates, and join with a world community of future-focused thinkers.
Unlock tomorrow’s traits immediately: learn extra, subscribe to our e-newsletter, and change into a part of the NextTech neighborhood at NextTech-news.com

Share. Facebook Twitter Pinterest LinkedIn Tumblr Email
NextTech
  • Website

Related Posts

Beijing Completes Rocket Road, China’s First Shared Testing and R&D Manufacturing Base for Industrial Spaceflight

January 18, 2026

Korea Growth Financial institution Launches ₩7.5 Trillion “Nationwide Progress Fund” to Again Deep Tech and AI Ventures – KoreaTechDesk

January 18, 2026

First-of-Its-Form Breakthrough: GLM-Picture, Educated Solely on Chinese language-Made Chips, Tops Hugging Face Trending

January 18, 2026
Add A Comment
Leave A Reply Cancel Reply

Economy News

His Excellency Sheikh Nahyan bin Mubarak Al Nahyan crowns winner of Longines Grand Prix class at FBMA Worldwide Present Leaping Cup 2026

By NextTechJanuary 18, 2026

His Excellency Sheikh Nahyan bin Mubarak Al Nahyan, Minister of Tolerance and Coexistence, attended the…

Mario’s Tiny Odyssey Has Him Conquering the Apple Watch

January 18, 2026

Beijing Completes Rocket Road, China’s First Shared Testing and R&D Manufacturing Base for Industrial Spaceflight

January 18, 2026
Top Trending

His Excellency Sheikh Nahyan bin Mubarak Al Nahyan crowns winner of Longines Grand Prix class at FBMA Worldwide Present Leaping Cup 2026

By NextTechJanuary 18, 2026

His Excellency Sheikh Nahyan bin Mubarak Al Nahyan, Minister of Tolerance and…

Mario’s Tiny Odyssey Has Him Conquering the Apple Watch

By NextTechJanuary 18, 2026

Tobi, a proficient coder, took on the daunting activity of becoming Nintendo’s…

Beijing Completes Rocket Road, China’s First Shared Testing and R&D Manufacturing Base for Industrial Spaceflight

By NextTechJanuary 18, 2026

Based on Yizhuang Holdings on January 14, Rocket Road, a flagship challenge…

Subscribe to News

Get the latest sports news from NewsSite about world, sports and politics.

NEXTTECH-LOGO
Facebook X (Twitter) Instagram YouTube

AI & Machine Learning

Robotics & Automation

Space & Deep Tech

Web3 & Digital Economies

Climate & Sustainability Tech

Biotech & Future Health

Mobility & Smart Cities

Global Tech Pulse

Cybersecurity & Digital Rights

Future of Work & Education

Creator Economy & Culture

Trend Radar & Startup Watch

News By Region

Africa

Asia

Europe

Middle East

North America

Oceania

South America

2025 © NextTech-News. All Rights Reserved
  • About Us
  • Contact Us
  • Privacy Policy
  • Terms Of Service
  • Advertise With Us
  • Write For Us
  • Submit Article & Press Release

Type above and press Enter to search. Press Esc to cancel.

Subscribe For Latest Updates

Sign up to best of Tech news, informed analysis and opinions on what matters to you.

Invalid email address
 We respect your inbox and never send spam. You can unsubscribe from our newsletter at any time.     
Thanks for subscribing!