TechflierTechflierTechflier
  • Home
  • News
  • Features
  • Spotlight
  • About
    • Mission
    • Services
    • Contact
  • Newsletter
  • Shop
    • All Items
    • By Category
      • Bags
      • T-Shirts
    • Cart
Search
© 2025 Techflier. All Rights Reserved.
Reading: Moonshot AI launches Kimi K3, the world’s largest open-weights model
Share
Font ResizerAa
TechflierTechflier
Font ResizerAa
  • Home
  • News
  • Features
  • Spotlight
  • About
  • Newsletter
  • Shop
Search
  • Home
  • News
  • Features
  • Spotlight
  • About
    • Mission
    • Services
    • Contact
  • Newsletter
  • Shop
    • All Items
    • By Category
    • Cart
Have an existing account? Sign In
Follow US
© 2025 Techflier. All Rights Reserved.
News

Moonshot AI launches Kimi K3, the world’s largest open-weights model

The 2.8 trillion-parameter model tops coding benchmarks and challenges the notion that US AI companies are years ahead of China.

Techflier Staff
Last updated: July 17, 2026 3:02 pm
Techflier
Share
SHARE

Moonshot AI has launched Kimi K3, a 2.8 trillion-parameter model that is now the world’s largest open-weights AI system. The model has already topped the Frontend Code Arena leaderboard, beating Anthropic’s Claude Fable 5.

Kimi K3 uses a Mixture-of-Experts architecture with a 1 million-token context window. The model targets long-running coding sessions, repository-scale analysis, and agent workflows requiring sustained reasoning. Moonshot released two versions: K3 Max for chat and reasoning, and K3 Swarm Max for orchestrating multiple AI agents working in parallel.

The model’s open weights will be released on July 27. Arena CEO Anastasios Angelopoulos called it “the single biggest release of the year,” suggesting it may represent the moment China has surpassed the US in AI model capabilities. Moonshot claims Kimi K3 is within striking distance of GPT-5.6 Sol in several key benchmarks.

The launch intensifies the open-weight AI race. DeepSeek, Alibaba’s Qwen, and Thinking Machines have all released competitive open models in recent months, driving down inference costs and accelerating global AI adoption.

UAE spreads its 5GW AI campus across many sites
Protein-testing lab Adaptyv banks €34M for agent-run experiments
Guardio hits $1.1B value as Wiz co-founder joins its round
SaaS obituary aside, Linear commands a $2.5B valuation
Euno banks $23M so agents stop guessing at company data
TAGGED:AI benchmarkChinese AIcodingKimi K3large language modelMoonshot AIopen source
SOURCES:SiliconAngleTechStartups
Share This Article
Facebook Copy Link Print
Previous Article Hyundai Motor seeks full ownership of Boston Dynamics
Next Article TSMC adds $100B to Arizona fab investment after strong quarter

Get Some Gear

 

 

 

 

Quick Links

  • News
  • Features
  • Spotlight
  • Newsletter
  • Store

About Techflier

  • About Techflier
  • Services
  • Contact Us
  • Privacy
  • Legal

Indices

TechflierTechflier
Follow US
© 2026 Techflier. All Rights Reserved.
Welcome Back!

Sign in to your account

Username or Email Address
Password

Lost your password?