Anonymous request-token comparisons from Opus 4.6 and Opus 4.7
(news.ycombinator.com)
91.
92.
Opus 4.7 to 4.6 Inflation is ~45%
(news.ycombinator.com)
94.
‘Tokenmaxxing’ is making developers less productive than they think
(techcrunch.com)
95.
“Tokenmaxxing” is making developers less productive than they think
(techcrunch.com)
96.
Measuring Claude 4.7's tokenizer costs
(news.ycombinator.com)
97.
Claude Opus 4.7 costs 20–30% more per session
(news.ycombinator.com)
98.
Are the costs of AI agents also rising exponentially? (2025)
(news.ycombinator.com)
99.
Reid Hoffman weighs in on the ‘tokenmaxxing’ debate
(techcrunch.com)
100.
101.
Why Some Companies Say AI ‘Tokenmaxxing’ Is Key to Survival
(feeds.content.dowjones.io)
102.
103.
104.
AI companies are tightening token limits. The last one to blink may win
(feeds.feedburner.com)
105.
106.
Agent Reading Test
(news.ycombinator.com)
107.
Qwen-3.6-Plus is the first model to break 1T tokens processed in a day
(news.ycombinator.com)
108.
Caveman: Why use many token when few token do trick
(news.ycombinator.com)
109.
A Claude Code skill that makes Claude talk like a caveman, cutting token use
(news.ycombinator.com)
110.
Talk like caveman
(news.ycombinator.com)
111.
Show HN: sllm – Split a GPU node with other developers, unlimited tokens
(news.ycombinator.com)
112.
Device code phishing attacks surge 37x as new kits spread online
(bleepingcomputer.com)
113.
114.
AI companies charge you 60% more based on your language, BPE tokens
(news.ycombinator.com)
115.
116.
New EvilTokens service fuels Microsoft device code phishing attacks
(bleepingcomputer.com)
117.
Claude Code users hitting usage limits 'way faster than expected'
(news.ycombinator.com)
118.
Anthropic: Claude Code users hitting usage limits 'way faster than expected'
(news.ycombinator.com)
119.
The AI industry loves token inflation. Your company shouldn’t
(feeds.feedburner.com)
120.
Universal Claude.md – cut Claude output tokens
(news.ycombinator.com)