Interfaze: A new model architecture built for high accuracy at scale
(news.ycombinator.com)
1.
2.
How Sakana trained a 7B model to orchestrate GPT, Claude and Gemini LLMs
(venturebeat.com)
3.
Computer Use is 45x more expensive than structured APIs
(news.ycombinator.com)
4.
Computer Use Is 45x More Expensive Than Structured APIs
(news.ycombinator.com)
5.
6.
He asked AI to count carbs 27000 times. It couldn't give the same answer twice
(news.ycombinator.com)
7.
We decreased our LLM costs with Opus
(news.ycombinator.com)
8.
An update on recent Claude Code quality reports
(news.ycombinator.com)
9.
Claude Token Counter, now with model comparisons
(news.ycombinator.com)
10.
Two Months After I Gave an AI $100 and No Instructions
(news.ycombinator.com)
12.
Emotion Concepts and Their Function in a Large Language Model
(news.ycombinator.com)
13.
Emotion concepts and their function in a large language model
(news.ycombinator.com)
14.
1M context is now generally available for Opus 4.6 and Sonnet 4.6
(news.ycombinator.com)
15.
Anthropic upgrades Claude’s memory to attract AI switchers
(theverge.com)
16.
Qwen3.5 122B and 35B models offer Sonnet 4.5 performance on local computers
(news.ycombinator.com)
17.
What Claude Code chooses
(news.ycombinator.com)
18.
What Claude Code Chooses
(news.ycombinator.com)
21.
Claude Sonnet 4.6
(news.ycombinator.com)
22.
Trying out Gemini 3 Pro with audio transcription and a new pelican benchmark
(news.ycombinator.com)
24.
Claude Haiku 4.5
(news.ycombinator.com)
26.
27.
Anthropic launches new version of scaled-down ‘Haiku’ model
(techcrunch.com)
28.
Claude Sonnet 4.5 is Anthropic's safest AI model yet
(engadget.com)
29.