Chinese models dominate OpenRouter: exceeding 3 trillion tokens

Published on 2026-02-20 18:16 ℹ️ LocalLLaMA 📰 Read the original source article →

Modelli cinesi dominano OpenRouter: superati i 3 trilioni di token

Exponential growth of Chinese models on OpenRouter

The OpenRouter platform is witnessing a remarkable increase in the adoption of large language models (LLMs) developed in China. An unprecedented event was recorded this week: a model exceeded the threshold of 3 trillion tokens processed in just seven days.

Trillion token threshold exceeded

Moreover, for the first time, more than one model exceeds one trillion tokens in a single week. This data marks a significant change compared to what was observed even just a month ago, when only Grok reached similar volumes.

Changing balances

The rise of Chinese models highlights a shift in market balances, with performance that, according to OpenRouter data, exceeds that of models developed in the United States.

AI-Radar Takeaway

The OpenRouter platform is experiencing a surge in the use of language models of Chinese origin. For the first time, a model exceeds 3 trillion tokens processed in a week, and multiple models exceed one trillion, marking a shift from the dominance of US models.

🤖 Ask AI about this

Want to dive deeper? Read the full article from the source:

📖 READ THE ORIGINAL ARTICLE

💻 Need GPU Cloud Infrastructure?

For running LLM inference, training models, or testing hardware configurations, check out this platform:

⚡

RunPod GPU Cloud Platform

Flexible GPU cloud with pay-per-second billing. Deploy instantly with Docker support, auto-scaling, and a wide selection of GPU types from RTX 4090 to H100.

✓ No commitments ✓ Instant deployment ✓ Production-ready

🔗 This is an affiliate link - we may earn a commission at no extra cost to you.

💬 Comments (0)

🔒 Log in or register to comment on articles.

No comments yet. Be the first to comment!

🔍 Continue Exploring

SECTION

AI-Radar LLM On-Premise

Complete guide to running AI models locally: hardware, stack, privacy, and reference architectures.

→

Chinese models dominate OpenRouter: exceeding 3 trillion tokens

Exponential growth of Chinese models on OpenRouter

Trillion token threshold exceeded

Changing balances

💻 Need GPU Cloud Infrastructure?

💬 Comments (0)

🔍 Continue Exploring

Explore LLM On-Premise

Bias Beneath the Tone: Empirical Characterization of Tone Bias in LLM-Driven UX Systems

The Secret of LLM Models: Uncovering How Tokenizers Affect Their Performance

MiniMax M2.7 on OpenRouter: 204,800 token context window