Everything Horizon Labs in AI

Claude Sonnet 5 test over seven days of dense pro use. Writing, code, agents: what actually changes and who should upgrade.

Claude vs ChatGPT vs Gemini: we tested the three flagships (Sonnet 5, GPT-5, Gemini 3 Pro) across three real pro use cases. Verdict by profile.

ChatGPT Plus at $20 a month: we ran it a full month in a real pro workflow against the free plan. Deep Research, GPT-5, clear verdict.

Marlin, Sakana AI's agent, runs for 8 hours straight and returns a full strategy. We put it to the test: real capabilities, limits, and verdict.