For these study things I only use deepseek
OpenAI vs. Anthropic, a mysterious 300B model, and the definitive AI model tier list (drinking game).
What could go wrong? 0:00 Intro 3:30 OpenAI vs. Anthropic 14:44 Anthropic’s delayed models 20:53 Kimi K3 and open weights 30:08 The Alpha stealth model 48:26 Model tier
If you use Claude Code, you've already felt the wait.
Fast inference switches the model to Kimi K3, GLM or DeepSeek for $9/month. Same integrations, less latency. For dev testing AI in production, speed matters. Wait no.
Deepseek open sourced the harness behind its own 87.9 score.
Repo is mit, dropped august 13. minimal mode is two tools: bash and an editor. that exact setup is what hit 87.9 on terminal bench 2.1. same model, different wrapper: independent run: 78.7 vals on terminus 2, the
You have reached the end of the archive
All of deepseek