The article recounts a sensational three‑stage saga of AI agents infiltrating OpenAI, culminating in the Astra model that can run continuously for days, consume 75,000 tokens in 26 minutes, and outperform previous models on complex visual‑code benchmarks.
The author measures the token footprint of 223 MCP tools across three servers, discovers that only 44 are actually invoked, resulting in a 76% idle rate and a hidden cost of 36,000 tokens per conversation, and proposes scope‑based configuration to eliminate waste.
With GPT‑5.6 and newer models handling most routine coding steps, many previously essential Skills become redundant, leading to context bloat, token waste, and security risks; the article proposes a three‑category framework to keep only lightweight, high‑value Skills and outlines concrete pruning criteria.
In June, Coinbase CEO Brian Armstrong revealed an internal AI cost‑optimization program that cut the company's AI dollar spend by almost 50% while token consumption kept growing exponentially, achieved through five concrete measures involving model defaults, intelligent routing, cache reuse, context trimming, and transparent usage monitoring.
Peter Steinberger disclosed that his project consumed 603 billion tokens and cost over $1.3 million in a single month, with OpenAI covering the fee, sparking a debate about token‑driven AI development, automation pipelines built with Codex, and the sustainability of the token‑maxxing trend.
The article walks through configuring a custom model in Claude Code, installing the claude‑tap plugin, launching the tool, sending the message “Hello”, and then dissecting the resulting request to reveal token counts, latency, tool list, system prompts, message payload, and a lingering cache issue.
China’s AI models logged 4.69 trillion tokens in a week, overtaking the US, while OpenAI plans to add 3,500 jobs to reach 8,000 staff, Musk announced a terawatt‑scale space compute project, MiniMax M2.5 topped global usage, Tencent reshuffled its AI Lab, and Xiaomi launched three new models, highlighting a shift from parameter‑centric competition to real‑world deployment and cost efficiency.
In March 2026, OpenRouter recorded 9.55 trillion tokens consumed weekly, with Chinese models occupying six of the top‑10 slots, Qwen surpassing 1 billion downloads, and cost advantages that let domestic LLMs outpace U.S. counterparts in both performance and price.
The author shares a 2025 personal usage report for Cursor, revealing a Top 25% activity rank with 1.41 B tokens consumed, and compares this with heavy‑use data from Cursor team members and industry leaders who spend up to 9.1 B tokens, highlighting model preferences and usage patterns.
An in‑depth analysis of 2025 H1 OpenRouter token usage shows explosive growth in Q1, highlights Google Gemini’s market dominance, reveals diverse long‑tail demand across domains, and examines shifting API preferences, offering key insights into the evolving landscape of large‑model services.
