TGViewer
Discover • Tech News Discover • Tech News @perplexity · 1.44M subscribers
Post #758 153K
🇨🇳 Chinese Open Source LLMs Capture 30% of the Market

Open-weight models are rapidly reshaping the market, according to a new report by OpenRouter and a16z, based on the analysis of over 100 trillion tokens of real LLM usage.

OpenRouter is a "unified API" for hundreds of AI models, allowing developers to access LLMs from different providers without setting up separate integrations.


While at the end of 2024, Chinese open-source models accounted for an average of just 1.2% of weekly tokens, by 2025 their share had risen to nearly 30%, thanks to releases from DeepSeek, Qwen, and other products.

Moreover, this growth hasn't been limited to a "hype wave": the share of open models has remained stable, indicating real-world adoption.

According to OpenRouter, open-weight models have become a true alternative to closed AI: they are cheaper, more flexible, customizable, and especially attractive for tasks where high stability isn't required, and openness is important.
  • ❤ 202
  • 👍 28
  • 🔥 22
  • 🤔 11
  • 👏 10
  • 👀 9
  • 🤷‍♀ 4
  • 🙏 3
  • 🗿 3
  • 😢 2
  • 🥴 1
More from @perplexity
  1. Oct 7, 2026🔴 Nano Banana 2.1 Is Now on @GPT4Telegrambot Google's updated model is much better at get…
  2. Oct 6, 2026🔥 Teen Gets Stranded on a Cliff After Letting AI Plan His Hike A 16-year-old hiker set ou…
  3. Oct 5, 2026📌 ChatGPT Will Start Adding Hidden Watermarks to Text in the EU ChatGPT and Codex will st…
  4. Oct 4, 2026🤖 74% of Physical Work in the US Can Already Be Done by Robots This work accounts for 34%…
  5. Oct 3, 2026🤖🤖 Mecha Fights Are Coming to Tokyo. Startup Skeletonics is opening a permanent venue wh…
  6. Oct 1, 2026🎮 Halloween Kicks Off in Fortnite Fortnite's annual Halloween event, Fortnitemares, has b…
Threads Profile ViewerView any public Threads profile without an account.Open ThreadLook →Writing with AI? Make it sound human.Metric37 rewrites AI drafts so they read naturally. Free AI detector, 1,500 words free.Try Metric37 →