DeepSeek-V4-Flash Tops OpenRouter Weekly Usage with 7.22 Trillion Tokens
DeepSeek-V4-Flash has topped the weekly usage charts on OpenRouter, a major AI model aggregator, with over 7.22 trillion tokens processed between July 27 and August 2. Additionally, the OpenCode platform reported a single-day peak of 8 trillion tokens for the model, driven by both free trials and paid developer usage. This milestone highlights the rapid global adoption of cost-effective Chinese AI models, which currently dominate the top spots on OpenRouter. It signals a shift in developer preferences toward highly efficient, budget-friendly alternatives to established Western models. DeepSeek-V4-Flash is a Mixture-of-Experts (MoE) model featuring 284 billion total parameters, with 13 billion activated per token, and supports a large 1-million-token context window. Out of the 8 trillion tokens processed in a single day on OpenCode, 5 trillion were from free trials while 3 trillion were paid.
## BACKGROUND
OpenRouter is an API aggregator platform that provides developers access to over 400 large language models (LLMs) through a single unified endpoint. A token is the basic unit of text (such as a word or subword) that LLMs use to process and generate language. Mixture-of-Experts (MoE) is an architecture that activates only a subset of a model's parameters for each task, significantly reducing computational costs while maintaining high performance.