Users praise OpenRouter's 100 trillion token study on open versus closed models because the released data is viewed as interesting, meaningful, and useful.
Based on 3 visible X reactions from 6 accounts; directional sample.
Ask a question below.
Published answers will appear here.
@beautyisfinite @AnjneyMidha @OpenRouter @MaikaThoughts @alexatallah @cclark by all means this data is interesting/meaningful/useful, im not questioning this, rather im speculating on what these numbers might look like for ALL the world's usage across every provider, rather than the subset of openrouter usage that we actually have data for
If the past year demonstrated that agentic inference is viable at scale, the next will focus on operational excellence: measuring real task completion, reducing variance under distribution shifts, and aligning model behavior with the practical demands of productionscale workloads" makes sense
@AnjneyMidha @OpenRouter @MaikaThoughts @alexatallah @cclark This is awesome!! Thanks for sharing, and for making most of the data available here https://openrouter.ai/rankings
ppl are increasingly asking for high quality empirical data on open vs closed frontier model usage a fairly rigorous paper on this topic is our recent @OpenRouter 100 trillion token study, led by @MaikaThoughts @alexatallah @cclark and team it is up on arxiv https://x.com/AnjneyMidha/status/2075790365362155694/photo/1
Users praise OpenRouter's 100 trillion token study on open versus closed models because the released data is viewed as interesting, meaningful, and useful.
Based on 3 visible X reactions from 6 accounts; directional sample.
Ask a question below.
Published answers will appear here.