I kinda dislike that in most benches the Chinese models are not included
@Valuy, one thing that gets undersold in these ‘challengers catching up’ stories: Anthropic and OpenAI pricing has dropped dramatically because of Chinese model pressure. So even if you stick with the incumbents, you’re benefiting from the competition. The real disruption isn’t which model wins — it’s that API costs across the board are trending toward commodity. For early-stage builders that’s actually a better story than picking a winner. tracks the pricing curve across major providers if you want the historical view.
GLM 5.2 on openrouter costs $0.93/M input and $3/M output. claude opus 4.8 costs $5/M input and $25/M output. claude is only a “good deal” if you consider the subsidized plans for private users that let you use 10x or more the price you paid for.
That being said, I’m not a fan of AI at all.
Each new model of OpenAI’s has had increased costs. Where are you getting this trending down thing from?
Anthropic models have stayed largely the same.
I like this article downplaying the model by saying this is only a “mini deepseek moment in a developer centric sense” - as if developers aren’t the main profitable use case for LLMs and the ones responsible for the model selection in products that reach end users
If the model is good enough for a developer and the cost is right, then that’s the model that you’re going to be using whether you notice or not.
