# State of AI: An Empirical 100 Trillion Token Study with OpenRouter

arXiv: https://arxiv.org/abs/2601.10088 · HTML: https://mlsanity.com/paper/2601.10088/

## Key facts

- Cited by 18 papers in the map (data through 14 Oct 2026).
- Builds on 6 of its 8 references that are in the map.
- First posted on arXiv 15 Jan 2026.

## Builds on (most cited first, 6 of 6)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022)
- [Chatbot Arena: An Open Platform for Evaluating LLMs by Human Preference](https://mlsanity.com/paper/2403.04132/) (2024)
- [WildChat: 1M ChatGPT Interaction Logs in the Wild](https://mlsanity.com/paper/2405.01470/) (2024)
- [Anthropic Economic Index report: Uneven geographic and enterprise AI adoption](https://mlsanity.com/paper/2511.15080/) (2025)

## Built on by (most cited first, 18 of 18)

- [AgencyBench: Benchmarking the Frontiers of Autonomous Agents in 1M-Token Real-World Contexts](https://mlsanity.com/paper/2601.11044/) (2026)
- [How are AI agents used? Evidence from 177,000 MCP tools](https://mlsanity.com/paper/2603.23802/) (2026)
- [OmniOPD: Logit-Free On-Policy Distillation via Speculative Verification](https://mlsanity.com/paper/2606.01476/) (2026)
- [Don't be so Stief! Learning KV Cache low-rank approximation over the Stiefel manifold](https://mlsanity.com/paper/2601.21686/) (2026)
- [FlashAccel: Leveraging High-Bandwidth Flash (HBF) for High-Throughput LLM Inference](https://mlsanity.com/paper/2607.10186/) (2026)
- [Understanding Usage and Engagement in AI-Powered Scientific Research Tools: The Asta Interaction Dataset](https://mlsanity.com/paper/2602.23335/) (2026)
- [Quantifying Self-Preservation Bias in Large Language Models](https://mlsanity.com/paper/2604.02174/) (2026)
- [The Cognitive Divergence: AI Context Windows, Human Attention Decline, and the Delegation Feedback Loop](https://mlsanity.com/paper/2603.26707/) (2026)
- [NanoCP: Request-Level Dynamic Context Parallelism for Data-Expert Parallel Decoding](https://mlsanity.com/paper/2605.21100/) (2026)
- [Personalization Meets Safety:Mechanisms,Risks,and Mitigations in Personalized LLMs](https://mlsanity.com/paper/2606.09038/) (2026)
- [An Experimental Design Approach to Evaluating Agentic AI's Autonomous Model Discovery](https://mlsanity.com/paper/2607.06413/) (2026)
- [MAPS: Memory-Aware Predictive Scheduling Framework for Large Language Model Serving](https://mlsanity.com/paper/2609.15359/) (2026)
- [Persistent Billable State: Denial-of-Wallet Attacks and Defenses in Tool-Calling LLM Agents](https://mlsanity.com/paper/2609.28585/) (2026)
- [When Fancy Eviction Fails: Rethinking Cache Replacement For LLM Prefix Reuse](https://mlsanity.com/paper/2609.28870/) (2026)
- [Beyond Memory Construction: Rethinking Memory Access for LLM-based Conversational Agents](https://mlsanity.com/paper/2609.33226/) (2026)
- [Has LLM Screening Performance Stalled in Software Engineering Systematic Reviews?](https://mlsanity.com/paper/2610.10633/) (2026)
- [The Wealth of Agents: Scaling Agentic AI Across Developers](https://mlsanity.com/paper/2610.20518/) (2026)
- [Price of Anarchy of First-Price Auctions in Agentic Markets](https://mlsanity.com/paper/2610.20587/) (2026)

## Abstract (arXiv)

The past year has marked a turning point in the evolution and real-world use of large language models (LLMs). With the release of the first widely adopted reasoning model, o1, on December 5th, 2024, the field shifted from single-pass pattern generation to multi-step deliberation inference, accelerating deployment, experimentation, and new classes of applications. As this shift unfolded at a rapid pace, our empirical understanding of how these models have actually been used in practice has lagged behind. In this work, we leverage the OpenRouter platform, which is an AI inference provider across a wide variety of LLMs, to analyze over 100 trillion tokens of real-world LLM interactions across tasks, geographies, and time. In our empirical study, we observe substantial adoption of open-weight models, the outsized popularity of creative roleplay (beyond just the productivity tasks many assume dominate) and coding assistance categories, plus the rise of agentic inference. Furthermore, our retention analysis identifies foundational cohorts: early users whose engagement persists far longer than later cohorts. We term this phenomenon the Cinderella "Glass Slipper" effect. These findings underscore that the way developers and end-users engage with LLMs "in the wild" is complex and multifaceted. We discuss implications for model builders, AI developers, and infrastructure providers, and outline how a data-driven understanding of usage can inform better design and deployment of LLM systems.