Models
Model
The underlying AI system (like GPT-4, Claude, or Gemini) that processes your requests.
Also written as models
- Messages
- 2,453
- Members using it
- 199
- First seen
- Nov 17, 2025
What it means
The underlying large language model that powers AI tools—the trained neural network that processes prompts and generates responses. Different models (GPT-4, Claude Opus, Gemini, local models) have different strengths, costs, context windows, and capabilities.
How it's used here
Members constantly compare models for specific tasks, test outputs across multiple models simultaneously, and debate which model works best for coding, research, creative work, or reasoning. The community tracks model releases closely and shares performance insights.
Seen in the wild
“Now just orchestrate that part which finds deficiencies the other model's output had and you've almost nailed SaaS automation building down.”
“I actually don’t arrange by book. I get the agent to rip out the mental models, principles and skill based knowledge and archive it that way”
“But really wish they would standardise reasoning token windows so you could compare how models perform with e.g. 25k reasoning token budgets”